A team of Seattle high schoolers will compete against underwater robotics teams from around the world during the MATE ROV World Championships in Canada next month.

Triton Robotics is making its third straight trip to the international event in St. John’s, Newfoundland. The independent team is comprised of seven 11th graders who all attend Seattle Academy.

MATE (Marine Advanced Technology Education) ROV, which takes place June 25-27, challenges teams to complete underwater mission tasks drawn from real-world ocean science problems. This year’s tasks include mapping cold-water coral ecosystems, deploying ocean observatory instrumentation, modeling offshore wind turbines, and operating profiling floats beneath sea ice — a notoriously difficult environment for conventional monitoring equipment.

The 2026 competition is themed around two United Nations science initiatives focused on ocean sustainability and cryospheric research.

Triton is bringing two custom-built systems to the competition. Njord, a remotely operated vehicle, will handle the wave tank and flume tank challenges, navigating strong currents, low visibility, and precision manipulation tasks. Skadi, an autonomous vertical profiling float, will operate in the National Research Council’s ice tank, diving beneath sea ice that conventional ocean-monitoring floats can’t reach.

Triton also built TritonOS, custom onboard software that handles depth-hold stabilization and lets the pilot instantly flip Njord’s controls so the rear manipulator operates as intuitively as the front arm — something no commercial system could do. The team added specialized tools for the mission tasks, including computer vision for invasive crab identification, a photogrammetry pipeline for iceberg measurement, and pneumatic grippers calibrated using tomatoes until they could grasp delicate cold-water coral without crushing it.

“This season has been about iteration — testing, breaking things, understanding why they failed, and rebuilding them better,” said Tenzin Larkin, the team’s co-CEO.

Triton Robotics is fully student-led and unaffiliated with any school, according to a news release. Members oversee engineering design, software development, budgeting, testing, documentation, and pool operations; mentors provide workshop and safety oversight. The team completed approximately 30 deep-water pool tests this season with zero safety incidents.

Team members include Griffin Fisher, Tenzin Larkin, Thomas Gust, Simon Hajduk, Theo Lipson, Miles Lipson, and Emi Enoki.

“Our mission is to help the environment through engineering,” said Gust, the team’s other co-CEO.

Update: A Triton Tech team from Edmonds College north of Seattle will also compete in the same event. The team — which is raising funds — is competing for the third consecutive year, and is in the “Pioneer” class against colleges from around the globe. In its inaugural year, a last-minute team built an ROV from PVC pipe and placed fifth in its category. Last year, the team traveled to Michigan and finished ninth in the world.

Two teenagers from Bellevue, Wash., took home a combined $125,000 at the world’s largest high school science competition this month — one for a low-cost filter that could help save Puget Sound’s salmon, the other for an AI-powered device that expands access to music therapy.

Lakshmi Agrawal, a senior at Interlake High School, and Anusha Arora, a sophomore who also attends Interlake, were among more than 1,700 students from roughly 60 countries who competed at the Regeneron International Science and Engineering Fair in Phoenix.

The annual competition, run by the Society for Science, is the world’s largest science and engineering contest for high schoolers and handed out more than $7 million in awards this year.

Agrawal, 18, won the Regeneron Young Scientist Award and a $75,000 prize for developing a low-cost, biodegradable sponge that filters a tire-derived chemical linked to the mass die-offs of coho salmon in Puget Sound-area streams. The chemical, 6PPD-quinone, has been identified as a primary culprit in kills that wipe out up to 80% of returning adult coho in some urban waterways before they can spawn.

To tackle the problem, Agrawal turned to an unlikely raw material: waste fibers from the jute plant. In lab tests, her sponge-like filters removed up to 80% of the pollutant from water containing tire particles, and also captured heavy metals and other contaminants. Compared to existing filtration alternatives, her approach required 85% less energy to produce and cut costs by roughly 98%.

Agrawal is headed to MIT in the fall to study chemistry and chemical engineering.

Learn more about her project here.

Arora, 15, took home the F. Thomson Leighton and Bonnie Berger Family Prize for STEM Excellence and $50,000 for building a portable music therapy platform that uses artificial intelligence to generate personalized music in real time based on a listener’s emotional state.

The device reads biometric signals through finger sensors and runs them through a suite of 11 AI models to detect emotions and compose adaptive music on the fly. In testing, users showed measurable reductions in stress and anxiety and stayed more engaged with therapy sessions than with conventional approaches.

Arora designed the platform to address a gap she identified in mental health care — music therapy is a clinically recognized treatment, but cost, provider shortages and spotty insurance coverage put it out of reach for most people.

Learn more about her project here.

Maya Ajmera, president and CEO of Society for Science, said she never fails to be inspired by the students who compete in the fair.

“They come from different backgrounds, different disciplines, and different corners of the world, and they are taking on some of our most urgent challenges with rigor, imagination, and determination,” Ajmera said in a statement. “At a moment when bold thinking is needed most, they are proof of what’s possible. I couldn’t be more optimistic about the future.”

The first CEO of the Fund for Science and Technology, the $3.1 billion foundation responsible for giving away a large portion of the late Microsoft co-founder Paul Allen’s fortune, has stepped down nine months after its public launch.

Dr. Lynda Stuart, a physician-scientist who previously led the Institute for Protein Design at the University of Washington, announced her departure on Thursday, saying the foundation will have “different leadership” for its next phase. No reason was given for the change.

The Seattle-based foundation launched publicly in August with a mandate to deploy at least $500 million over four years in bioscience, the environment, and AI for good.

It has awarded at least $30 million in grants to institutions including Fred Hutchinson Cancer Center, Seattle Children’s, Benaroya Research Institute, and Scripps Oceanography.

Stuart’s last day was May 8, according to a spokesperson for the Allen estate, who said Stuart made the decision to step down. Chief Financial and Operations Officer Liz Carey and Chief Programs Officer Marc Malandro will lead the organization in the interim, with the board starting a CEO search in the future.

The change follows broader shifts in the Allen philanthropic ecosystem. Earlier this year, the foundation’s shift toward applied uses of AI and a new proposal-based funding process for the Allen Institute for AI contributed to the departure of several top researchers — including former Ai2 CEO Ali Farhadi, who joined Microsoft with at least nine other colleagues.

It also comes at a pivotal moment more broadly for the Allen estate, which is led by Paul Allen’s sister, Jody Allen, as the estate works to sell his professional sports franchises.

• The sale of the Portland Trail Blazers was approved by the NBA Board of Governors in March, in a two-part deal valued at a reported $4.25 billion.
• The Seattle Seahawks went on the market in February, weeks after winning the Super Bowl, and are expected to sell for upward of $9 billion, according to ESPN.

Allen directed that all proceeds from the sale of his sports teams and other assets would go to philanthropy, and FFST is expected to oversee a large portion of those proceeds.

Jody Allen chairs the FFST board. The board also includes Allan Jones, president emeritus of the Allen Institute; Tom Daniel, a UW professor emeritus; and Nancy Peretsman, a managing director at investment bank Allen & Company.

Former Microsoft CEO Steve Ballmer, who was on the board when FFST launched in August, is no longer listed as a member. Ballmer, who served on the board from its formation in 2022, completed his service in March, the spokesperson for the Allen estate confirmed.

In her LinkedIn post, Stuart alluded to the challenging federal funding environment for science, saying FFST’s early investments “provided hope when scientists didn’t have any” and “preserved key programs that would have otherwise been dismantled.” She said the foundation will remain grounded in Seattle, continuing to contribute to Allen’s legacy.

Stuart, reached via direct message, referred to her post and directed further inquiries to the foundation. As for what’s next for her, she wrote in her public post: “Science is a global enterprise, and as a citizen of the world, who knows what that might be?”

Long before Homer wrote the Odyssey, Minoan seafarers were plying the trade routes of the Mediterranean and spinning stories of adventure — but when it comes to imagination on an outlandish scale, the late science-fiction author Vonda N. McIntyre’s tale about a transoceanic Minoan odyssey just might have awed even Homer.

McIntyre finished the manuscript for her final book, “The Curve of the World,” less than two weeks before she died of pancreatic cancer in 2019. Since then, a team of writers and editors assembled by Clarion West — the Seattle-based literary nonprofit that McIntyre founded in 1971 — has been working to get the novel in shape for publication.

That work is now complete. The book has made its debut, and Clarion West is celebrating with a virtual book launch party on Saturday. Update: You can replay the party on YouTube.

Nisi Shawl, an award-winning science-fiction and fantasy writer, admits to “fan-girling” during the editing process. “The sheer joy of the prose, the sensual array of delights that are offered, every bit of the way, the writing is just so pleasurable,” Shawl says in the latest episode of the Fiction Science podcast.

McIntyre was arguably best known for her novelizations of three “Star Trek” movies, plus two books based on the original TV series, a “Star Wars” novel and a totally original “Starfarers” book series. One of her award-winning novels, “The Moon and the Sun,” was adapted for a movie that was titled “The King’s Daughter” and released after her death.

“The Moon and the Sun” blended history, fantasy, science fiction and romance in an alternate-universe version of 17th-century France under the reign of Louis XIV. “The Curve of the World” follows a similar trajectory — but set in the Bronze Age, about 3,600 years ago.

In a sense, the tale is complementary to Homer’s Odyssey, which is getting the Hollywood treatment this summer in a movie starring Matt Damon as Odysseus. “It’s a side-by-side emergence of a literal hero’s journey and a literal heroine’s journey,” Shawl says.

McIntyre’s main character is a woman, not a man. She’s a diplomat, not a warrior. And her odyssey is the reverse of Odysseus’ homeward voyage. The long and winding trip begins at home, on the island of Crete, and goes far beyond the Mediterranean Sea to the Americas.

Along the way, the Minoan crew encounters Nordic pirates on the high seas, hunter-gatherers on North America’s Atlantic Coast and a bloodthirsty society in ancient Mesoamerica that keeps its mummified ruler around long after death. Perhaps the most outlandish stretch in McIntyre’s tale has the crew crossing over to the Pacific side of the Americas and sailing up to, of all places, Puget Sound.

The Minoans encounter a strange new world where the inhabitants live in longhouses, weave the fur of woolly dogs into their blankets, and feast on cedar-planked salmon. But it’s no utopia: The visitors become entangled in intertribal conflicts as well as natural threats that are all too familiar to the Pacific Northwest’s modern-day residents.

Although the Minoans were skilled travelers and traders, they weren’t as skilled as they’re portrayed to be in McIntyre’s novel. “While I’d love to believe some intrepid Minoans crossed the Pacific using the stars, unfortunately the archaeological evidence for anything beyond the Mediterranean remains extremely limited,” Alessandro Berio, a Brazilian archaeoastronomer who has studied the Minoans extensively, said in an email.

“The Minoans, generally considered Europe’s first advanced maritime civilization, established the earliest large-scale trade network in the Mediterranean. Archaeological material has been recovered throughout Egypt, the Levant and Anatolia,” Berio said. But he added that clear evidence for Mediterranean navigators venturing beyond the Mediterranean itself “only appears later, during the Phoenician period.”

That’s the difference between the sort of history that Berio studies and the fiction-flavored history that was McIntyre’s specialty. Shawl is a practitioner of that type of history as well: Shawl’s best-known novel, “Everfair,” describes a steampunk world in which socialists and missionaries unite to foster an independent nation in the Belgian Congo at the dawn of the 20th century.

“I’m starting to come to the conclusion that ‘alternate history’ is probably a misnomer, because sometimes it’s sort of like a veering away from the history we know, right?” Shawl says. “Sometimes it’s a hidden history, something that happened within the interstices of the history that we know. And sometimes it’s just straight-up a different version, a different perspective.”

McIntyre was known for highlighting different perspectives when it came to gender and sexual diversity. There’s plenty of that in “The Curve of the World.” For example, one of the characters smoothly switches between acting as a man or a woman, depending on the circumstance.

Shawl found the novel’s main character, a trader and emissary named Iakinthu, to have a particularly interesting perspective on the cultures she encounters. “She was not being, like, traditionally aggressive, but she was not giving way, either,” Shawl says. “She was very much, ‘You got to do you, but when it comes to our interactions, you do me.’ I really want to study how Vonda portrayed her doing that, because I think it will be helpful in modern times.”

In the book’s acknowledgments, McIntyre provided a cautionary note about real Minoan history vs. her alternate “Idaean” history: “Do not try to match the Idaean timeline to ours because your head will explode,” she wrote.

“My head did explode,” Shawl admits. “My head exploded because I was trying to hold together all the different layers of reality that she was carving up and serving. Other people have told me of earlier incidents where they were like, ‘Is this how it really could have happened? Is this what did happen? What is Vonda doing?’

“Well, she’s just being great, that’s all,” Shawl said.

Clarion West is presenting a virtual book launch party for “The Curve of the World” at 11 a.m. PT Saturday. Follow the link from this webpage to RSVP and get the Zoom link for the event. Shawl and two other book editors — Debbie Notkin and L. Timmel Duchamp — will read from the book and discuss the project. Kath Wilham also participated in the editing process.

Clarion West is also presenting a “Curve of the World” reading and conversation with Shawl and Theo Downes-Le Guin — the literary executor for author Ursula K. Le Guin’s estate — at Elliott Bay Book Company in Seattle at 7 p.m. PT on July 16. Follow the link from this webpage to RSVP and get in-person or virtual tickets.

My co-host for the Fiction Science podcast is Dominica Phetteplace, an award-winning writer who is a graduate of the Clarion West Writers Workshop and lives in San Francisco. To learn more about Phetteplace, visit her website, DominicaPhetteplace.com.

Fiction Science is included in FeedSpot’s 100 Best Sci-Fi Podcasts. Check out the original version of this report on Cosmic Log to get sci-fi reading recommendations from Shawl, and stay tuned for future episodes of the Fiction Science podcast via Apple, Spotify, Player.fm, Pocket Casts and Podchaser. If you like Fiction Science, please rate the podcast and subscribe to get alerts for future episodes.

University of Washington physicist David Hertzog can’t wait to find out how hundreds of researchers who worked on a geeky project known as the Muon g-2 Collaboration will react when they hear they’ve each won thousands of dollars for that work.

The money is coming from this year’s $3 million Breakthrough Prize for fundamental physics, which was awarded tonight during a gala ceremony in Los Angeles. Hertzog and his colleagues decided that the prize should be divided equally among everyone who was an author on research papers relating to the decades-long series of muon experiments.

“There are students who were in and out of this thing — two years or less,” Hertzog said. “They’re going to be shocked out of their lives about something they did a long time ago that they don’t remember doing. They’re going to get a phone call or email from the Breakthrough people, and they’re going to go, ‘What!?’ That’s kind of fun.”

Hertzog said payments will go to about 400 researchers who were involved in the Muon g-2 experiments at Fermilab in Illinois and at the Brookhaven National Laboratory in New York. In addition, the prize honors the early role played by Europe’s CERN research center, going back to 1959. Fifteen authors involved in CERN’s experiments will also get a share of the money.

Although the precise math hasn’t yet been worked out, dividing $3 million among 400 people would give each recipient $7,500. “That’s nothing to throw around if you’re a student or a young postdoc,” Hertzog said.

A big moment for the muon

Israeli tech investor Yuri Milner and his wife, Julia Milner, established the Breakthrough Prize in 2012 to recognize achievements in fundamental physics, mathematics and the life sciences. They also wanted to add some Hollywood pizazz to the public perception of scientists, going so far as to spread out a red carpet for the “Oscars of Science.” The host for tonight’s ceremony was James Corden, and the guest list included Robert Downey Jr., Anne Hathaway, Paris Hilton and Lionel Richie. Bill Gates and his girlfriend Paula Hurd showed up as well.

The $3 million Breakthrough Prize is the world’s richest scientific award, outdoing the roughly $1.2 million prize given to Nobel laureates. More than $344 million has been handed out since the creation of the prize program. Past winners from the University of Washington include physicists Eric Adelberger, Lukasz Fidkowski, Jens Gundlach and Blayne Heckel, plus biochemist David Baker.

This year’s prize in fundamental physics touches on a long-running effort to reconcile experimental findings with one of history’s most successful scientific theories: the Standard Model of particle physics. The theory lays out a framework for classifying and understanding a menagerie of subatomic particles — including the muon, which is similar to the electron but 207 times heavier.

The Standard Model predicts the various properties of the muon. One such property is the strength and orientation of the muon’s magnetic field, known as its magnetic moment. The theory’s simplest formulation calls for the value of the muon’s magnetic moment, represented in equations by the letter g, to be equal to 2.

Few things in particle physics are that simple, however. Experimental tests measured the g-factor to be slightly more than 2, and that discrepancy became the focus of the Muon g-2 (pronounced “mew-on gee-minus-two”) experiments.

If there was a confirmed mismatch between the Standard Model and experimental results, that could open the door to new physics. For example, perhaps whole new sets of subatomic particles not predicted by theory had somehow eluded direct observation. So, physicists across the globe marshaled their forces to determine the value of g, either to fill in the gap between experiment and theory or to zero in on a new frontier in physics.

Over the years, physicists have been conducting increasingly fine-tuned experimental runs using powerful magnets at CERN, Brookhaven and Fermilab. Hertzog has been in on the quest since Brookhaven joined in, about 30 years ago, and he was part of the team in 2013 when the experiment’s massive main magnet was moved from Brookhaven to Fermilab.

Each run narrowed the uncertainty surrounding the precise value of g. The crowning achievement came from Fermilab’s version of the experiment in 2025.

“We set the goal at 140 parts per billion, and we got 127 parts per billion,” Hertzog said. “When we wrote the proposal, we were ambitious as we could get in our minds, because we wanted to get people to take us on. Then we just blew away all the systematic errors, better than we expected. And then new ones came along, which caused us to have a little bit of a struggle.”

At the same time, other physicists were wrestling with theoretical models. They factored in the ever-so-subtle effects of particles popping in and out of the quantum foam that’s thought to make up the fabric of spacetime at its smallest scale. Last year, one of the models came up with a range of theoretical values for g that overlapped with the Muon g-2 Collaboration’s range of experimental values.

That led some to claim that there was no discrepancy after all. “A famous particle physics experiment has ended not with a bang, but a whimper,” Science magazine reported. But once again, few things in particle physics are that simple. Hertzog insisted that reports of the muon mystery’s death have been greatly exaggerated.

“I just throw up my hands, because after 30-some years of working on this, it’s a little disappointing that it’s not clear,” he said. “Not only has the number that they recommended shifted, but the certainty of their number got way wider. The uncertainty on the theory recommendation is actually pretty big. It’s shifted, but it’s also pretty large.”

Hertzog said the Breakthrough Prize recognizes a scientific quest that’s still in progress. “This story is not finished,” he said. “The story is really about the extraordinary achievement of the precision of this delicate measurement which probes nature to such a deep, deep level.”

Will there ever be a definitive answer to the muon mystery?

“We don’t know it yet, but it’s knowable, as opposed to walking out into a vast cloud of ambiguity,” Hertzog said. “So, I think we will find out in a couple of years where that finally lands. … Who knows whether that’ll lead us to another chapter in this business. But I’m confident that we’ll know it.”

A big night for breakthroughs

The Muon g-2 Collaboration’s Breakthrough Prize was awarded to hundreds of researchers from 31 institutions in seven countries, but just four team members were selected to take the stage for tonight’s award ceremony. Hertzog was joined by Chris Polly from Fermilab, William Morse from Brookhaven, and Lee Roberts from Brookhaven and Boston University.

A special lifetime prize for fundamental physics went to David Gross, a theorist at the Kavli Institute of Theoretical Physics at the University of California at Santa Barbara. Gross won a share of the 2004 Nobel Prize in Physics for filling gaps in the Standard Model relating to the strong nuclear force. More recently, he helped write a landmark 40-year national plan for particle physics.

Three prizes were given in the life sciences:

• Jean Bennett, Katherine High and Albert Maguire of the University of Pennsylvania were recognized for developing a therapy for inherited retinal degeneration that became the first gene therapy approved by the Food and Drug Administration for a genetic disease.
• Another prize went to Stuart Orkin, a physician at Harvard Medical School and Boston Children’s Hospital; and to Swee Lay Thein at the National Heart, Lung and Blood Institute for elucidating the mechanism driving the switch from fetal to adult hemoglobin and validating it as a therapeutic target for sickle-cell disease and beta-thalassemia.
• Rosa Rademakers of the Mayo Clinic and Bryan Traynor of the National Institute on Aging won the third prize in the life sciences for discovering the most common genetic cause of ALS and frontotemporal dementia.

Frank Merle of the Institut des Hautes Études Scientifiques in Paris was awarded this year’s prize in mathematics for achieving breakthroughs in nonlinear evolution equations. His work could have implications from aeronautical engineering and safety to astrophysics.

For his part, Hertzog doesn’t intend to rest on his laurels. Even as the Muon g-2 Collaboration is winding down, he has joined the team for another particle physics experiment called PIONEER. That experiment will probe inconsistencies between the Standard Model and observations of pion decay. As was the case with the Muon g-2 experiments, there’s a chance that PIONEER could point the way to physics beyond the Standard Model.

“This is a stock market golden opportunity,” Hertzog said. “That’s how I look at it.”

The Breakthrough Prize website has the full list of this year’s honorees, including the winners of New Horizons Prizes for early-career physicists and mathematicians, Maryam Mirzakhani New Frontiers Prizes for women mathematicians and the inaugural Vera Rubin New Frontiers Prize for women physicists. The recorded awards show is due to air at noon PT on April 26 via YouTube.

Hertzog and University of Bern physicist Martin Hoferichter review the results of the Muon g-2 experiments in a preprint paper titled “The Anomalous Magnetic Moment of the Muon: Status and Perspectives.”

Updates as of 9:40 a.m. April 21: We’ve corrected a reference to the CERN phase of the Muon g-2 experiments to note that 15 living researchers from that era shared in the Breakthrough Prize. We’ve also gone to a more precise description of Yuri Milner’s nationality. Milner was born in Russia but renounced his Russian citizenship in 2022 and has been an Israeli citizen since 1999.

Microsoft and NASA say they’re applying artificial intelligence to a challenge that has become increasingly urgent: how to cope with flooding and other disasters driven by extreme weather.

The result of their efforts is Hydrology Copilot, a set of AI agents aimed at making hydrological data easier to access and analyze. The platform is built on the foundation that was established for NASA Earth Copilot, a cloud-based AI tool that can sift through petabytes of Earth science data.

Hydrology is the scientific study of Earth’s water cycle, which encompasses precipitation, runoff, evaporation and the movement of water through rivers, lakes and soil. It’s not just an academic exercise: Hydrologic insights are put to use in fields ranging from agriculture to forestry to urban development.

“NASA has long produced advanced hydrology and land-surface datasets, powering breakthroughs in drought early-warning systems, environmental planning and environmental research,” Juan Carlos López, a senior solution specialist at Microsoft who focuses on space and AI, wrote in a blog post. “Yet despite their value, these datasets and the specialized tools required to navigate and interpret them remain difficult to access for many who could benefit most.”

That’s where Hydrology Copilot comes in: Powered by Microsoft Azure OpenAI Service and Microsoft Foundry, the platform lets researchers and others query NASA’s data using straightforward questions — for example, “Which regions may be facing elevated flood risk?”

Hydrology Copilot simplifies access to one of NASA’s most advanced hydrology datasets, the North American Land Data Assimilation System Version 3. The dataset integrates satellite measurements with computer models to present a continuously updated, high-resolution, continental-scale view of the water cycle.

Insights gleaned from the database can help planners figure out how to improve drought monitoring, agricultural planning, water resource management, flood risk assessment and emergency preparedness. Recent flooding in Western Washington, fueled by a succession of atmospheric rivers, shows how important it can be to gain such insights — and how useful Hydrology Copilot can be.

“The goal of this project is to provide the tools that can enable local officials, city planners and emergency responders to more easily understand weather patterns and better prepare for the types of hydrological events we are seeing now in the Pacific Northwest and around the world, and will likely continue to see in the future,” a Microsoft spokesperson told GeekWire in an emailed statement.

Hydrology Copilot is still under development, and is being used primarily by researchers for now. Microsoft’s Azure AI team can provide further information about the platform. To take a test spin through more publicly accessible hydrology datasets, check out King County’s Hydrologic Information Center and the interactive map provided by the National Water Prediction Service.

Mount Rainier is not getting ready to erupt, but scientists at the Pacific Northwest Seismic Network might blow their stacks over having to respond to a rumor about increased tremor activity at the Washington state volcano.

The Daily Mail reported this week that the mountain was sending up “a flurry of strange signals” and raising concern “that something inside the volcano might be shifting.”

“This towering stratovolcano looms over more than 3.3 million people across the Seattle-Tacoma metro area, threatening to cripple entire communities with ashfall, flooding, and catastrophic mudflows if it erupts,” The Daily Mail reported in dramatic fashion in a story that went viral online.

PNSN was forced to respond on Wednesday with a detailed blog post debunking the story, explaining that a likely buildup of ice on the antenna of a Rainier seismic station was causing radio interference misread as seismic activity.

PNSN said the station at St. Andrews Rock (STAR) is “one of the last remaining old analog sites on the mountain and uses very low-power radio transmission to send data down to a receiver site.” The system, located just west of the summit, is susceptible to weather-related interference and PNSN said recent stormy weather likely caused “rime ice buildup.”

The blog post contains recent seismograms and further explanation about how the station’s typical signals are affected by glacier slips, rock falls and high winds. PNSN said any unusual seismic activity such as significant earthquakes or volcanic tremor would show up on numerous other stations that are on or near the volcano and would be detected and reported by PNSN within a short time.

“We realized that what they were seeing was one, just one, of our many seismic stations that was sending out essentially static, just noise, instead of actual data,” PNSN Director Harold Tobin told KING 5. “And that’s the whole problem, and the thing that people think is tremor at the volcano, it’s just some static.”

PNSN is operated cooperatively by the University of Washington, the University of Oregon, and the U.S. Geological Survey (USGS)’s Seattle Field Office to monitor earthquake and volcanic activity across the Pacific Northwest.

The last minor eruption at Mount Rainier was recorded in 1884.

The last minor eruption at PNSN was Wednesday in response to The Daily Mail.

“Unfortunately, sloppy journalism by non-scientists who don’t understand seismology nor check with those that do can generate confusion in the public and more work for those who need to correct the clearly incorrect information,” PNSN’s blog post read. “It is no wonder that some publications are only considered tabloids and should never be believed.”

Previously:

• Seattle shakes again as Mariners grand slam triggers another big burst of seismic activity

Creating a virtual brain may sound like a science-fiction nightmare, but for neuroscientists in Japan and at Seattle’s Allen Institute, it’s a big step toward a long-held dream.

They say their mouse-cortex simulation, run on one of the world’s fastest supercomputers, could eventually open the way to understanding the mechanisms behind maladies such as Alzheimer’s disease and epilepsy — and perhaps unraveling the mysteries of consciousness.

“This shows the door is open,” Allen Institute investigator Anton Arkhipov said today in a news release. “It’s a technical milestone giving us confidence that much larger models are not only possible, but achievable with precision and scale.”

Arkhipov and his colleagues describe the project in a research paper being presented this week in St. Louis during the SC25 conference on high-performance computing. The simulation models the activity of a whole mouse cortex, encompassing nearly 10 million neurons connected by 26 billion synapses.

To create the simulation, researchers fed data from the Allen Cell Types Database and the Allen Connectivity Atlas into Supercomputer Fugaku, a computing cluster developed by Fujitsu and Japan’s RIKEN Center for Computational Science. Fugaku is capable of executing more than 400 quadrillion operations per second, or 400 petaflops.

The massive data set was translated into a 3-D model using the Allen Institute’s Brain Modeling ToolKit. A simulation program called Neulite brought the data to life as virtual neurons that interact with each other like living brain cells.

Scientists ran the program in different scenarios, including an experiment that used the full-scale Fugaku configuration to model the entire mouse cortex.

“In our simulation, each neuron is modeled as a large tree of interacting compartments — hundreds of compartments per neuron,” Arkhipov said in comments emailed to GeekWire. “That is, we are capturing some sub-cellular structures and dynamics within each neuron.”

During the full-scale simulation, it took no more than 32 seconds to simulate one second of real-time activity in a living mouse brain. “This level of performance — 32 times slower than real time — is quite impressive for a system of this size and complexity,” Arkhipov said. “It is not uncommon to see a factor of thousands of times slower for such very detailed simulations (even much smaller than ours).”

The researchers acknowledge that much more work is needed to turn their simulation into a model capable of tracing the progress of a neurological disease. For example, the model doesn’t reflect brain plasticity — that is, the brain’s ability to rewire its own connections.

“If we want to mention something specific besides plasticity, then one aspect that is missing is the effects of neuromodulators, and the other is that we currently do not have a very detailed representation of sensory inputs in our whole-cortex simulations,” Arkhipov said. “For all of these, we need much more data than currently available to make much better models, although some approximations or hypotheses could be implemented and tested now that we have a working whole-cortex simulation.”

Arkhipov said the project’s long-term goal is to simulate an entire brain, not just the cortex. “There’s a distinction between whole-cortex and whole-brain,” he pointed out. “The mouse cortex (and our model of it) contains about 10 million neurons, whereas the whole mouse brain contains about 70 million neurons.”

A human-brain simulation would require an even greater leap. The human cortex alone contains not just 10 million neurons, but 21 billion.

The good news is that a sufficiently powerful supercomputer might be up to the task. “Our work shows that very detailed microscopic-level simulations of larger brains may be possible sooner than previously expected,” Arkhipov said. “The results suggest that a simulation of the whole monkey brain (such as that of a macaque monkey with 6 billion neurons) can fit on the full-scale Fugaku system.”

Arkhipov said it was important to point out that creating a brain model on a supercomputer “does not mean that such a model is complete or accurate.”

“Here we are talking about technical feasibility of simulations, and it looks like such simulations even at the scale of the monkey brain are now within reach,” he said. “But to make such simulations biologically realistic, much more experimental data production and model building work would need to happen.”

Rin Kuriyama and Kaaya Akira of the University of Electro-Communications in Tokyo are the principal authors of the paper presented at SC25, titled “Microscopic-Level Mouse Whole Cortex Simulation Composed of 9 Million Biophysical Neurons and 26 Billion Synapses on the Supercomputer Fugaku.” In addition to Arkhipov, authors from the Allen Institute include Laura Green, Beatriz Herrera and Kael Dai. The study’s other authors are Tadashi Yamazaki and Mari Iura of the University of Electro-Communications; Gilles Gouaillardet and Asako Terasawa of the Research Organization for Information Science and Technology in Hyogo, Japan; Taira Kobayashi of Yamaguchi University; and Jun Igarashi of the RIKEN Center for Computational Science.

Dozens of Seattle biotech companies are using artificial intelligence to design new medical treatments. But at a conference of industry leaders and investors this week, scientists delivered a nuanced message: AI holds enormous promise, but expectations need to stay grounded in reality.

The trade association Life Science Washington and investment firm Madrona hosted the one-day forum in downtown Seattle delving into biotech, pharma and AI.

“There is some over hype on the breadth and the depth of the capabilities of some of these AI models,” said Jamie Lazarovits, CEO and co-founder of Archon Biosciences. Researchers need “to be really cautious” when drawing conclusions from the data generated by the models, he said.

And yet there’s a lot to be excited about.

“What was science fiction 15 years ago is now reality,” said Erik Procko, chief scientist for Cyrus Biotechnology. “So yes, there’s been hype. But there’s still enormous potential, and sometimes the progress that is being made is just dizzying.”

Lazarovits and Procko were part of a panel including four Seattle startups harnessing AI. Each company is tackling different challenges in drug development:

• Archon, a company that emerged from stealth more one year ago with $20 million in financing, is using AI to design proprietary protein structures, known as Antibody Cages or AbCs, that are intended to help antibodies bind to target cells and avoid other cells.
• Cyrus is designing drugs with a focus on identifying and removing areas that will trigger an immune-system response — a challenge called immunogenicity. The 10-year-old company has raised $36.6 million according to PitchBook.
• Outpace Bio, a startup founded in 2021 that has raised $200 million, is engineering proteins that aim to bolster T-cell therapies for solid tumors, which constitute 90% of cancers. Tumors are discouragingly good at deflecting current T-cell treatments, which often stop working within a month.
• Talus Bioscience, which launched in 2020 and has nearly $20 million, is targeting transcription factors — proteins that are part of the “regulome” that turns genes off and on. The company is targeting transcription factors that activate genes that drive specific cancers.

Beyond discussing their own work, the panelists identified key principles for how AI should — and shouldn’t — be used in biotech research:

Augmenting researchers, not replacing them

Marc Lajoie, co-founder and CEO of Outpace, compared AI tools to the robotic exoskeleton donned by Ripley and others in the sci-fi film Aliens to move heavy cargo — and battle the ET Xenomorph Queen.

“It makes the researcher better,” he said. “We’re not trying to replace the researcher.”

Models must meet reality

Lazarovits noted that while AI can generate exciting leads and information, it doesn’t mean much until it has been tested in actual experiments with cells and organisms.

“Whenever we try to adopt new models, new AI methods, you can become incredibly excited by having this in silico validation,” he said. “But the reality is, is what do you actually validate in the wet lab?”

The real bottleneck: clinical trials

AI is great for engineering new therapies, but the most costly, laborious part of the drug development process is seeing how they work in patients.

“The most impactful place for AI to really change the game for drug development would be to make smaller, better powered clinical studies,” said Lajoie. The way to do that, he added, was coming up with better drug candidates that perform multiple functions.

Still searching for AI’s breakthrough moment

Procko is still waiting for AI to go further in spurring advancements in biotech and pharmacology.

“AI is fantastic these days for predicting, say, a protein structure, but for creating new drugs it hasn’t yet found its killer app,” said Procko. “What is it that AI is letting us do now to make new drugs that was simply impossible to make before? How is it being a game changer?”

The question captures a central tension discussed at the panel and conference: while AI has transformed how Seattle’s biotech companies approach drug design, the industry is still navigating the gap between computational promise and clinical proof.

Editor’s note: Story updated Oct. 12, 2025 to correct the reference to “in silico validation” made by Jamie Lazarovits.

The groundbreaking immunology research behind two Nobel Prizes awarded this week traces back to a largely forgotten Seattle-area biotech startup that was backed by Microsoft’s co-founders.

Darwin Molecular launched in 1992, and Mary Brunkow and Fred Ramsdell — who on Monday won Nobel Awards in Physiology or Medicine — joined two years later. Bill Gates and Paul Allen both invested in the Bothell, Wash.-based company, whose team included the star biologist Lee Hood.

At the time, the Human Genome Project was just getting up to speed in sequencing the DNA code that programs humans, and Darwin aimed to clone individual genes from within that master plan. Hood recounted the startup’s efforts at an event Tuesday at Institute for Systems Biology (ISB) that celebrated Brunkow, who is a senior program manager at the Seattle-based nonprofit research organization.

When the company started, there was no complete genome to work from, said Hood, co-founder of ISB. “You had to do it from scratch, and it was an extremely challenging idea,” he added. “And there was enormous skepticism.”

Brunkow recounted her excitement over landing a role at a biotech firm instead of academia.

“We made a great team at this small startup company where we were given the freedom to try different things,” she said. “And it was a really exciting time.”

Brunkow tackled a deadly autoimmune disease in mice, initiating a genetics program to breed the mice and isolate the mutation. The scientists sifted through DNA sequences and ultimately found the gene, which they named FOXP3, and pinpointed the genetic error that caused it to malfunction.

Ramsdell’s team focused on the cell biology and immune system operations in the mutant mice.

After isolating the gene, Brunkow and her team were able to explore new hypotheses about its cellular function, ultimately uncovering new insights into how the immune system regulates and suppresses its own responses.

Brunkow, Ramsdell and Shimon Sakaguchi of Japan’s Osaka University were jointly awarded the Nobel Prize for helping identify regulatory T cells, dubbed the “immune system’s security guards.”

The cells help prevent the body’s immune system from attacking the wrong targets — which if mismanaged can lead to autoimmune diseases such as rheumatoid arthritis, multiple sclerosis, lupus and many others.

The researchers also showed that a mutation in the human version of the FOXP3 gene caused the rare, life-threatening autoimmune disease IPEX.

Brunkow and Ramsdell demonstrated that FOXP3 is “a hallmark” of regulatory T cells, said Jim Heath, president of ISB. “Since that time, this discovery has had an impact in virtually every aspect of human health and disease.”

Ramsdell is the co-founder and chair of the scientific advisory board for Sonoma Biotherapeutics, a company with R&D facilities in Seattle.

Darwin was acquired in 1996 by London-based Chiroscience Group, which three years later merged with another British company, Celltech. The biotech firm ultimately shutdown the R&D operations in Washington state in 2004 — which is when Brunkow and Ramsdell left the company.

“I would still be working on that project because it really was an awesome time and an awesome team,” Brunkow said. “And we knew we were doing something really important in the field.”

RELATED: Seattle scientists Mary Brunkow and Fred Ramsdell win Nobel Prize for immunology research

Scientists say an artificial intelligence program that they compare to ChatGPT has helped them create one of the most detailed maps of the mouse brain to date, with 1,300 regions and subregions marked on the map.

Some of those subregions have never been charted before — and the researchers say there’s more to come. “I think there are already indications that we can go beyond what we see now,” said Bosiljka Tasic, director of molecular genetics at Seattle’s Allen Institute for Brain Science.

The mapping effort, led by researchers at the University of California at San Francisco and the Allen Institute, is detailed in a study published today in the journal Nature Communications.

“Our model is built on the same powerful technology as AI tools like ChatGPT,” senior author Reza Abbasi-Asl, a neuroscientist at UCSF, said in a news release. “Both are built on a ‘transformer’ network which excels at understanding context.”

That context could be important for treating neurological ailments, Tasic told GeekWire.

“Location is everything in the brain,” she said. “Defining the geography of the brain, and then defining all these regions and their functions, not only leads to better understanding, but also better ability to treat.”

More detailed maps of the brain’s cellular structure could lead to more targeted drug treatments that cause fewer side effects. “We always want to go toward better, more precise brain therapies, but in order to do that, you need to know where you need to interfere, what went wrong in what place, and what you need to fix,” Tasic said. “And if you don’t have the map, how are you going to know where it is?”

Mapping the brain’s neighborhoods

Brain-mapping efforts have typically relied on human interpretation of the brain’s anatomy, but scientists are getting better at identifying the location and function of millions of individual brain cells. They’re getting so much better at collecting huge masses of data that they need AI to help with the interpretation.

“We are at a point where we have amazing experimental technology, so next-generation sequencing is completely revolutionized,” Tasic said. “Our way to define cell types — the fact that you can measure thousands of genes per cell, and define cells that are similar as a cell type — has transformed biology.”

The availability of software that can deal with such high-dimensional data is making this “an amazing time for a neuroscientist,” she said.

The key to the newly published study is an AI model called CellTransformer. The model sifts through huge sets of data about the locations and functions of brain cells, known as spatial transcriptomics data sets, to determine which cells belong in the same “neighborhood” of the brain.

CellTransformer analyzed spatial transcriptomics data about 9 million cells in more than 200 tissue sections that were taken from the brains of four individual mice. At first, researchers programmed the model to define the boundaries of 25 regions in the brain. Eventually, they raised the resolution to define 670 regions and subregions. At each level of resolution, CellTransformer’s brain maps matched what had been defined previously by human experts.

Then the dial was turned up to produce 1,300 regions and subregions. At that level, CellTransformer successfully replicated maps of cataloged regions of the brain. It also identified previously uncataloged, finer-grained subregions in areas of the brain that are currently poorly understood.

Tasic said the process was like going from a map that showed only continents, or only countries, to a map that showed states, cities and even the neighborhoods within cities.

“What we’re saying is, let’s take any cell and ask, ‘Who are the neighbors?’ And then, based on the commonality of the neighbors, call it a region,” she said. “Basically, that’s what CellTransformer did.”

Some of the previously uncharted subregions are in the midbrain reticular nucleus, which plays a complex role in processing sensory and motor information. Other newly identified subregions are in the superior colliculus, a part of the midbrain that processes sensory information and initiates eye, head and body movements to focus on objects of interest.

Focusing on new neuro-frontiers

Tasic said it’s possible to turn up the dial on CellTransformer’s algorithms to produce maps of the brain that are even more detailed. “Now, the question is, which ones are meaningful, in what way, and what do they represent biologically?” she said.

Another question has to do with what to call the newly characterized subregions. “Just imagine that you came to a new land, and you are seeing there is this and there is that. But now I need to name it. Now I need to see what else is around,” Tasic said. “We want to give meaningful, systematic names and also reference how it relates to older maps.”

Perhaps the biggest questions relate to how the newly published map, which is based on cell types, will line up with maps that trace connections between cells, or patterns of brain-cell activity. “I’m just hoping for more systematic data collection, more systematic data analysis, and more multimodal models — models that will not only measure gene expression and cell type, but connectivity and activity, and define brain regions based on all of those,” Tasic said.

Tasic said the AI-based techniques that were developed for mapping mouse brains are “absolutely extendable to the human brain,” but she doesn’t expect that to happen overnight.

“The limit is actually data collection,” she said. “The human brain is huge, so that’s one problem. … I don’t want to give any estimates, but it will take maybe a decade more to just collect [data about] the full human brain at the level of detail that we did for the mouse.”

UCSF researcher Alex Lee is the principal author of the Nature Communications study, titled “Data-Driven Fine-Grained Region Discovery in the Mouse Brain With Transformers.” Other authors include Alma Dubuc, Michael Kunst, Shenqin Lao, Nicholas Lusk, Lydia Ng, Hongkui Zeng, Bosiljka Tasic and Reza Abbasi-Asl.

Mary Brunkow of the Seattle-based Institute for Systems Biology (ISB) and Fred Ramsdell of Sonoma Biotherapeutics, which has offices in Seattle and South San Francisco, today were awarded the 2025 Nobel Prize in Physiology or Medicine for their research on the immune cells that prevent the human body from attacking itself.

Shimon Sakaguchi of Osaka University in Japan will also share the prize for their work identifying regulatory T cells, dubbed the “immune system’s security guards.”

“Their discoveries have been decisive for our understanding of how the immune system functions and why we do not all develop serious autoimmune diseases,” Olle Kämpe, chair of the Nobel Committee, said in a statement.

In 1995, Sakaguchi made the initial discovery of immune cells that protect the body from autoimmune diseases, but many in the field were skeptical of his conclusions.

Six years later, Brunkow and Ramsdell were working together at Celltech, a biotech company that previously had an R&D facility in Bothell, Wash. They identified a gene they named FOXP3 that, when damaged, caused severe autoimmune illness in mice. They also showed that a mutation in the human version of this gene caused the rare, life-threatening autoimmune disease IPEX.

The research done by the trio solidified the role of regulatory T cells, or Treg cells, and launched a new field in peripheral tolerance research, which is the study of how the body makes the distinction between dangerous invaders and its own cells. Work in this area has led to therapies for cancer and autoimmune diseases.

Brunkow earned her bachelor’s degree in Molecular and Cellular Biology from the University of Washington, and holds a master’s degree and PhD in Molecular Biology from Princeton University.

When the Nobel Committee tried to reach Brunkow today to inform her of the award, she ignored the call, the Associated Press reported.

“My phone rang and I saw a number from Sweden and thought: ‘That’s just, that’s spam of some sort,’” Brunkow told the AP.

“When I told Mary she won, she said, ‘Don’t be ridiculous,’” said her husband, Ross Colquhoun.

Brunkow worked at Celltech for a decade, until the United Kingdom-based company closed its Bothell facility in 2004. She has been a senior program manager at ISB for more than 16 years.

“ISB congratulates Dr. Brunkow on this extraordinary honor and celebrates her contributions to science and human health,” ISB President Jim Heath said via email.

ISB is a nonprofit studying cancer, aging, infectious disease, chronic illness and other health challenges. Renowned Seattle scientist Dr. Lee Hood co-founded the organization in 2000 for research into systems biology, an interdisciplinary field that tackles biology from an integrated perspective. ISB is an affiliate of the Providence health care system.

Ramsdell is based on Bainbridge Island, just west of Seattle. In 2019 he helped co-found Sonoma Biotherapeutics, which opened a new R&D center and office space on Seattle’s waterfront last year. The company is focused specifically on Treg cells. It has a partnership with Regeneron to co-develop cell therapies for Crohn’s disease, ulcerative colitis, and two undisclosed conditions.

“Like many great scientists, [Ramsdell] spent years working behind the scenes to identify the gene that caused devastating systemic autoimmune diseases in a little-known strain of mice,” Jeff Bluestone, Sonoma CEO and president said in a statement. “The discovery of the gene, FOXP3, changed our understanding of peripheral tolerance and led to a new field of immunotherapy.”

Ramsdell earned his PhD in microbiology and immunology from the University of California, Los Angeles and holds a bachelor’s degree in biochemistry and cell biology from the University of California, San Diego.

His career includes roles at Seattle biotech giants Immunex and Zymogenetics, and he spent six years at Celltech. Ramsdell was previously chief scientific officer at Sonoma, and is now chair of the company’s scientific advisory board.

The Seattle area is home to numerous companies and organizations working on autoimmune diseases, a list that includes but is not limited to:

• Adaptive Biotechnologies has done research on T-cell receptor genetic sequences associated with Crohn’s disease.
• Bothell’s Sana Biotechnology has a drug program pursuing allogeneic CAR-T cells that target autoimmune diseases and leukemia.
• Mozart Therapeutics, a company targeting celiac and conditions including gastrointestinal disorders such as Crohn’s disease, Ulcerative Colitis and others.
• Indapta, a biotech company with facilities in Seattle and Houston, is using its universal natural killer cell platform to develop therapies.
• Seattle’s Alpine Immune Sciences, which was acquired last year by Boston-based Vertex Pharmaceuticals.

The prize was announced today at the Nobel Assembly at the Karolinska Institutet in Stockholm. The trio of winners will share 11 million Swedish kronor, which equals about $1.17 million.

Last year, UW biochemist David Baker won a Nobel Prize in chemistry for his research and discoveries about the molecular structure of proteins.

Editor’s note: Story updated to correct that Fred Ramsdell is based on Bainbridge Island, Wash., and to add additional information on his career.

REDMOND, Wash. — About 5 minutes before each hour, and 5 minutes after it, Microsoft Teams gets busy with people joining meetings early or late. Yes, you know who you are.

Those predictable spikes in demand are an example of a larger infrastructure challenge facing the industry: either build data centers full of chips and servers that sit idle for much of the time, waiting to handle peak loads, or risk performance issues during moments of high demand.

Microsoft has come up with a different approach. The company said Tuesday that it has successfully prototyped new advances in the field of microfluidics that let its servers run hotter and faster — a technique known as overclocking — without the risk of melting down the silicon.

“We’ve got these really spiky workloads — I would dearly love to be able to overclock … because then we would need way fewer servers and deliver a better experience,” said Jim Kleewein, a Microsoft technical fellow and development director for core Office 365 services, recounting the challenge he originally brought to Microsoft’s silicon development teams.

Microfluidics overall is not a new concept. Instead of placing a cooling plate on top of a chip, the approach brings liquid coolant directly inside the processor itself through tiny channels. 

But Microsoft says it has developed a new approach that uses AI to customize the cooling system for the unique heat signatures of different chips running specific workloads.

The design, inspired by biology, ends up resembling the veins in a leaf, delivering coolant with greater precision to a chip’s hottest spots. The company says it’s up to three times more effective at removing heat than current cold plates, making overclocking more feasible.

‘Surface of the sun’

Some of the hot spots on a chip “could have the same heat flux as the surface of the sun when you look at the very small scale,” said Husam Alissa, Microsoft’s director of systems technology, referring to a technical measure of heat intensity, or density in a specific area. 

While the working prototype was demonstrated on a commercially available Intel Xeon chip, Microsoft says the innovation is part of a broader strategy to improve its entire hardware fleet. As a next step, the company is looking to incorporate the cooling technology into future versions of its own silicon, potentially including its Azure Cobalt chip and Maia AI accelerator.

Those first-party chips, unveiled in 2023, reflect Microsoft’s efforts to gain control over the final piece in its cloud platform. The company is competing to efficiently train and run cutting-edge AI models against Amazon, Google, and others, which are also making their own chips.

Microsoft still uses third-party chips, including industry-standard GPUs from Nvidia and other suppliers, but it says a more integrated approach is necessary to meet the challenges of the AI era. The goal isn’t to replace partners but to advance the broader ecosystem, said Rani Borkar, the corporate vice president who leads Azure hardware systems and infrastructure.

“The demands of Al are such that, frankly, hardware on its own cannot keep up,” Borkar said during a recent media briefing and tour of Microsoft’s Silicon Lab in Redmond. “We really are all about co-designing, co-optimizing every layer of the stack.”

Networking and steel

The company made two related announcements Tuesday morning:

• Microsoft is working with Corning and Heraeus to move to industrial scale production of its hollow core fiber technology, which transmits pulses of light through a hollow channel vs. a conventional solid glass core, letting data move faster and with lower latency.
• In an initiative focused on sustainability, Microsoft said it will procure “green steel” from partner Stegra for use in datacenter construction. The steel is produced in a plant designed to reduce carbon emissions by up to 95% vs. traditional steelmaking.

As for the newly announced microfluidics advances, Microsoft says it will work with partners and the broader tech community to make the technology an industry standard. 

Longer term, the company sees the cooling technology as a foundational step toward a revolutionary new type of chip architecture: 3D stacking. While stacking layers of silicon dramatically reduces latency by shortening the distance data has to travel, it has been largely unfeasible because of the immense challenge of removing heat from the inner layers.

Microfluidics could eventually solve that by allowing coolant to flow between each layer of the silicon brick, a breakthrough that Microsoft’s Kleewein believes could be transformative.

“That is where this could move from, ‘it’s an interesting change that the rest of the industry should adopt,’ to a ‘holy shit’ moment in the evolution of technology,” he said.

Our brains are wired to fill in perceptual gaps in what we see, whether it’s a lion hiding in the trees or the shapes hidden in an optical illusion — but how does that wiring work? Neuroscientists are zeroing in on how special kinds of brain cells help us see things that aren’t actually there.

Researchers from Seattle’s Allen Institute for Brain Science and the University of California at Berkeley traced the role played by the cells, known as IC-encoder neurons, in a study published today by the journal Nature Neuroscience.

“The goal of this project was to understand the neural basis of pattern completion, or filling in when you are dealt ambiguous or missing data in your vision,” said senior study author Hillel Adesnik, a neuroscientist at Berkeley.

Such research could help scientists understand how our brains create a complete picture of the world around us from the data that our senses provide. It could also eventually reveal how hallucinations arise, or point the way to better computer vision systems.

The researchers behind the newly published study conducted their experiments using mice. They showed the mice a series of diagrams — some of which used Pacman-like circles to create an illusory black bar, and others in which the bar was explicitly outlined.

The Allen Institute’s OpenScope Brain Observatory team recorded how the brains of the mice responded to the different diagrams, millisecond by millisecond, and then delivered the data to the Berkeley team.

“The observatory provided access to what was going on during illusions across the brain,” Jerome Lecoq, an associate investigator at the Allen Institute, told GeekWire in an email. “This allowed the experiment at Berkeley in Hillel Adesnik’s lab to dive deeper into local mechanics that implemented those illusions.”

The Berkeley researchers identified a small set of neurons that responded to the diagrams with the illusion, but not to the diagrams with the outlined bars. They also used laser-based techniques known as two-photon calcium imaging and holographic optogenetics to trace the flow of activity between those IC-encoders — that is, illusion contour encoders — and other regions of the visual cortex.

The last set of experiments added one more twist: The researchers showed the mice a blank gray screen, but activated the IC-encoder neurons while they were doing it. Then they watched for activity in other neurons in the visual cortex.

“When we looked at those downstream neurons, their activity patterns looked more similar to when there was an actual illusory contour on the screen than when there was nothing. So we were sort of generating this controlled hallucination,” Adesnik said. “We can’t say that it actually made it to the highest levels of perception, assuming mice have perception. … But we could say that at the neural level, we could see neural pattern completion.”

Based on their analysis of the data, the researchers concluded that the IC-encoder neurons help manage the process of filling in the gaps in visual perception.

“The representation of the illusion arises in higher visual areas first, and then gets fed back to the primary visual cortex; and when that information is fed back, it’s received by these IC-encoders in the primary visual cortex,” lead study co-author Hyeyoung Shin, who went from Berkeley to Seoul National University in 2023, said in a news release.

Adesnik said future experiments could address at least some of the many questions raised by the newly published study. “If you really want to gauge what the mouse experiences perceptually, you’d have to train them to do so. And people have done this on all sorts of species — mice, monkeys, dogs, cats,” he said. “So we can do that.”

Insights from such research could eventually be applied to the human condition.

“In certain diseases you have patterns of activity that emerge in your brain that are abnormal, and in schizophrenia these are related to object representations that pop up randomly,” Lecoq said. “If you don’t understand how those objects are formed and a collective set of cells work together to make those representations emerge, you’re not going to be able to treat it; so understanding which cells and in which layer this activity occurs is helpful.”

Lecoq said the Allen Institute’s role in the research is consistent with the vision that gave rise to the OpenScope Brain Observatory back in 2018.

“The astronomical observatory analogy here is the observatory doing the survey of the sky, while the lab focuses on a local relevant constellation to untangle what is going on,” he said.

“This is how we view our Brain Observatory in the future: We will provide brain-wide recordings of specific and important behaviors, defined by one or more laboratories. Then those laboratories can analyze those datasets and sometimes perform complementary experiments to uncover specific computational mechanisms in the brain,” Lecoq said. “I am anticipating more research in neuroscience will be performed this way in the future.”

Previously: OpenScope team stirs up a brainstorm

In addition to Shin, Adesnik and Lecoq, the authors of the Nature Neuroscience study, titled “Recurrent Pattern Completion Drives the Neocortical Representation of Sensory Inference,” include Mora Ogando, Lamiae Abdeladim, Uday Jagadisan, Severine Durand, Ben Hardcastle, Hannah Belski, Hannah Cabasco, Henry Loefler, Ahad Bawany, Josh Wilkes, Katrina Nguyen, Lucas Suarez, Tye Johnson, Warren Han, Ben Ouellette, Conor Grasso, Jackie Swapp, Vivian Ha, Ahrial Young, Shiella Caldejon, Ali Williford, Peter Groblewski, Shawn Olsen and Carly Kiselycznyk.

Credit for Kanizsa Triangle graphic: Fibonacci, CC BY-SA 3.0, via Wikimedia Commons.

The Laser Interferometer Gravitational-Wave Observatory, or LIGO, has already won its researchers a Nobel Prize — and now artificial intelligence is poised to take LIGO’s search for cosmic collisions to the next level.

Google DeepMind and the LIGO team say they’ve developed an AI tool called Deep Loop Shaping that has been shown to enhance the observatory’s ability to track gravitational waves — faint ripples in the fabric of spacetime that are thrown off by smash-ups involving black holes and massive neutron stars.

The researchers describe the technique in a proof-of-concept study published today by the journal Science. They hope to make Deep Loop Shaping part of routine operations at LIGO’s detectors in Louisiana and on the Hanford nuclear site in Washington state.

“Deep Loop Shaping is revolutionary, because it is able to reduce the noise level in the most unstable and most difficult feedback loop at LIGO,” lead author Jonas Buchli, a research scientist at Google DeepMind, told reporters.

The existence of gravitational waves was predicted by Albert Einstein a century ago, but they were first observed directly only in 2015, using LIGO’s twin 2.5-mile-long interferometers. That feat was recognized with the Nobel Prize in physics in 2017. Since then, the LIGO team has worked to increase the sensitivity of the detectors, but it’s not an easy task.

LIGO’s system of mirrors and precision laser beams have to register spacetime warps that amount to just one-10,000th the width of a proton. At that level of sensitivity, perturbations from distant earthquakes and ocean waves can jiggle the mirrors enough to make a difference in what’s registered by the detectors.

Researchers developed sophisticated techniques — including some that rely on AI — to keep the 88-pound mirrors still and cancel out the “noise” from such perturbations. But compensating for that noise introduces a different kind of disturbance.

“The hardest problem left is, how do you keep everything so still without disturbing your measurement?” said Rana Adhikari, a member of the LIGO team at Caltech. “That ‘controls noise’ has been bedeviling us for decades and decades.”

Adhikari compared the problem to trying to hold a mirror still with your bare hands. “If you try to keep it really still, your hands start to shake because you’re holding it tightly,” he said. “This method takes away the shaking.”

Google’s engineers worked with LIGO’s scientists to develop software that was trained on simulated gravitational-wave readings, using a process known as reinforcement learning.

“Basically, they were running dozens of simulated LIGOs in parallel,” Adhikari said in a news release. “You can think of the training as playing a game. You get points for reducing the noise, and dinged for increasing it. The successful ‘players’ keep going to try to win the game of LIGO. The result is beautiful — the algorithm works to suppress mirror noise.”

The proof-of-concept test results, based on an hour’s worth of LIGO data from the detector in Louisiana, showed that Deep Loop Shaping could quiet the motion of the mirrors 30 to 100 times better than traditional noise-reduction methods alone.

Study co-author Jan Harms, a professor at Italy’s Gran Sasso Science Institute, said the technique could open up a new frontier in astronomy.

“We are really excited about the potential to advance gravitational-wave science with Deep Loop Shaping,” he said. “More specifically, what we can now do is open a new frequency band for gravitational-wave observations, toward the low-frequency end.”

Harms said the opportunity was analogous to expanding the range of a telescope to take in infrared or X-rays as well as optical wavelengths. With greater sensitivity at lower frequencies, LIGO could do a better job of detecting the collisions of neutron stars or intermediate-mass black hole binaries. LIGO could also provide more advance warning of imminent cosmic collisions.

“What you can do is a pre-merger alert, so you can let people know [that] a minute from now, two neutron stars will merge,” Harms said. “And then, if you have just enough detectors online, you can even point to a specific patch in the sky and tell them, ‘Look there and wait for it.’”

With all the talk about AI hallucinations, should people be worried that Deep Loop Shaping could produce false data? “I think the question of ‘Will it misbehave after it’s running a year’ is a legitimate question, but we also worry about that for our classical methods, and so we monitor all of these things,” Adhikari said.

“This is a new area for us, so I think we’ll be learning as we go along, and we’ll be developing methods to veto any sort of misbehavior, not just from this system, but even our classical systems that do misbehave sometimes,” he said.

In the next phase of the rollout, the LIGO team plans to put Deep Loop Shaping through longer test runs that could go on for days or weeks at a time — in Louisiana, at Hanford and eventually at LIGO-India. “This week is the time when we’re going to start to have those discussions,” Adhikari said.

Deep Loop Shaping, and programs like it, could well become part of the standard engineering toolkit — not just for gravitational-wave detectors, but also for other applications where components have to be controlled with high precision. “This is about applications in aerospace, for example,” Harms said. “Navigation or manufacturing, or generally noise reduction in systems, or also civil engineering.”

The technology might even find its way into noise-canceling headphones. And Buchli said there may be still other applications that engineers haven’t yet thought of.

“I think once we send this out, hopefully some more people come to think, ‘Oh, actually, gee, I have this really hard control problem. I think I will try that,’” he said.

Buchli, Adhikari and Harms are among 30 authors of the Science study, “Improving Cosmological Reach of a Gravitational Wave Observatory Using Deep Loop Shaping.”

Yes, ChatGPT can draft a tricky email. But can a general-purpose AI assistant for the masses contribute to a big scientific breakthrough, without hallucinating or screwing up the research? The Allen Institute for AI (Ai2) isn’t holding its breath.

The Seattle-based institute’s new project, dubbed Asta, lets research scientists create and deploy specialized AI agents, working autonomously to help them tackle and solve some of the world’s toughest problems.

Ai2 bills Asta as a “bold initiative to accelerate science by building trustworthy and capable agentic assistants for researchers.” It’s also an open-source project, following the nonprofit institute’s ethos of making the digital plumbing of its projects available for others to analyze, adapt, and build upon.

There are three major components to Asta:

• Asta agents: AI assistants that can find papers, analyze data, summarize literature, generate hypotheses, and cite their sources.
• AstaBench: a benchmark suite for testing and ranking scientific agents on real-world problems.
• Asta Resources: a developer toolkit for building scientific agents using Ai2’s open-source tools and data.

With entry-level jobs in some industries already under pressure, Ai2 is careful to call Asta a collaborator for scientists, not a replacement. But most grad students would envy the kind of praise the institute is heaping on the project.

Ai2’s chief scientist, Dan Weld, said the inspiration came from the institute’s own need for AI tools that can execute complex plans, explaining how they came up with their ideas, and stay grounded in scientific evidence.

“That’s what Asta delivers,” Weld said in the announcement. “It’s not just another assistant but a collaborator designed to think like a scientist.”

The institute sees this open approach — where developers use Asta’s resources to build agents and Asta’s benchmark to test them — as a way to create “a flywheel of scientific improvement” for the broader community.

The news caps what will no doubt go down as one of the busiest months in Ai2 history, including the announcement of a new AI robotics initiative and a $152 million grant from Nvidia and the National Science Foundation to lead the creation of the future AI backbone for U.S. scientific research.

In their spare time, the team has also been settling into new offices in a mass-timber building on the north end of Seattle’s Lake Union. 

Founded in 2014 by Paul Allen, and funded largely by the late Microsoft co-founder’s estate, Ai2 is a non-profit research institute that builds open-source AI systems for the good of science and humanity. 

Ai2’s team of 225 people includes some of the top AI researchers in the world, many of them also affiliated with the University of Washington’s Allen School of Computer Science & Engineering. Ai2 is known for OLMo (open language models), Molmo (multimodal models), Tülu (instruction-tuned language models), and other open-source AI systems.

OpenAi, Anthropic, and other Silicon Valley startups have grabbed headlines for turning their proprietary AI models into widely used consumer and commercial products, while Ai2 has worked more quietly behind-the-scenes on foundational research and fully open-source tools for scientists and developers.

Asta is an outgrowth of Semantic Scholar, Ai2’s long-running project to create a massive research database that indexes more than 200 million scientific papers. Asta’s agents use this database through a dedicated “Scientific Corpus Tool,” an extension of the Semantic Scholar API.

Ai2 has been burnishing its open-source credentials under CEO Ali Farhadi, an entrepreneur and University of Washington professor who returned to Ai2 in 2023 after leading machine learning at Apple — which acquired his Ai2 spinout Xnor.ai for an estimated $200 million in 2020.

Farhadi says the goal is to hold AI to standards grounded in the scientific method: He said in a statement that Asta represents Ai2’s “bet on a future where AI doesn’t just keep up with science, it helps drive it forward.”

Is this the Allen Institute for AI’s breakout moment?

The U.S. National Science Foundation and chip powerhouse Nvidia will provide a total of $152 million in funding and infrastructure to the Seattle-based non-profit research institute to lead a new national project to build open AI models for scientific research.

The five-year initiative positions Ai2, as it’s known, as a key player in building the technological backbone of U.S. artificial intelligence, while expanding its sources of support beyond its historic reliance on grants and philanthropy from the estate of late Microsoft co-founder Paul Allen.

It marks NSF’s first major investment in AI software infrastructure for science.

The initiative “will supercharge the work we do at Ai2 and increase America’s ability to deliver breakthrough AI developments,” Ai2 CEO Ali Farhadi said in a statement. 

The project, Open Multimodal AI Infrastructure to Accelerate Science (OMAI), will develop a family of large, fully open, multimodal AI models trained on scientific literature and data. The goal is to accelerate breakthroughs in fields such as materials science, biology, and energy.

AI is becoming “not just a tool for research, but a catalyst for rethinking how discovery happens across every discipline,” said the OMAI project leader, Noah Smith, Ai2 senior director of natural language processing research and a computer science professor at the University of Washington’s Allen School of Computer Science & Engineering.

Consisting of $75 million from the NSF and $77 million from Nvidia, the initiative is part of the White House AI Action Plan to strengthen U.S. leadership in AI-driven research.

“The question isn’t if AI will transform science — it’s who will lead that transformation, and this investment helps ensure that the answer is America,” said Tess deBlanc-Knowles, special assistant for artificial intelligence at NSF, in a briefing with reporters this week. 

The project gives Ai2 a rare level of access to key AI hardware: Nvidia will supply its HGX B300 systems, built on the company’s new Blackwell Ultra architecture, along with AI Enterprise software to accelerate training and inference for the new models. 

Nvidia CEO Jensen Huang said in a statement about the project that the goal is “to generate limitless intelligence, making it America’s most powerful and renewable resource.”

The hardware is designed to handle massive datasets with high efficiency, speeding up the pace of scientific discovery, said Jack Wells, Nvidia’s director of higher education and research, during the briefing with reporters.

Founded in 2014 by Paul Allen, Ai2 has operated largely behind the scenes of the AI boom in recent years, focusing on research and open-source tools while for-profit giants such as OpenAI, Anthropic, Google DeepMind, and Meta have dominated headlines. 

Ai2’s work — from natural language models to scientific search engines — has earned respect in research circles, but the new federal and corporate backing stands to elevate its profile.

The project illustrates Washington state’s position “at the epicenter of the AI-powered scientific research revolution,” Sen. Maria Cantwell said in a statement, citing the efforts of the UW and Ai2 and saying that the initiative will “position our state’s researchers to drive AI breakthroughs in medicine, clean energy, and materials science.”

In leading the OMAI project, Ai2 will build on its experience creating OLMo and Molmo, two families of high-performance open AI models, along with open datasets such as Dolma. 

Unlike many leading AI systems that are fully proprietary or only partially open, Ai2’s models are released with the full weights, training data, code, and evaluation tools needed for researchers to inspect, adapt, and retrain them. The idea is to give scientists the transparency they need to recreate results, and trust the technology used in high-stakes research.

Also involved will be research teams from the University of Washington (led by Hanna Hajishirzi, also with Ai2), the University of Hawai’i at Hilo (Travis Mandel), the University of New Hampshire (Samuel Carton), and the University of New Mexico (Sarah Dreier).

Ai2 said the funding will primarily be used for computing power to train larger, more advanced models on the open foundations established with OLMo and Molmo. 

The OMAI models will be designed for scientific workflows, helping researchers process and visualize data, generate code for analysis, and identify patterns they might otherwise miss. The systems could also surface relevant findings from outside a researcher’s specialty, linking new insights to past discoveries to speed breakthroughs, according to Ai2. 

The project will roll out in stages, with datasets, code, and other resources released along the way. Ai2 expects the first major model to be available about 18 months into the five-year effort.

Reference to Nvidia HGX B300 corrected since publication.

Nathan Myhrvold, a Seattle tech titan who also studies titanosaurs and other denizens of the dinosaur era, realizes that “Jurassic World Rebirth” is science fiction, not a documentary — nevertheless, he has a few bones to pick with the filmmakers.

“There are some lines that it would be silly to cross, but they did anyway,” says Myhrvold, who was Microsoft’s first chief technology officer back in the 1990s and is currently the CEO of Bellevue, Wash.-based Intellectual Ventures.

Paleontology is one of Myhrvold’s many interests, and he’s a co-author of more than a dozen peer-reviewed papers on the subject. He was inspired to get into dinosaur research almost 30 years ago, when he visited a “Jurassic Park” movie set at the invitation of director Steven Spielberg. That visit led to connections with leading paleontologists.

“At that point in my life, I was interested in dinosaurs, but I’d never been professionally or seriously, in a scientific sense, into dinosaurs,” Myhrvold recalls. “So, the movie was a little bit instrumental in me, just as a way of meeting a bunch of those people.”

On the latest episode of the Fiction Science podcast, Myhrvold and University of Maryland paleontologist Thomas Holtz discuss how much scientists — and filmmakers — have learned about dinosaurs over the past three decades. And they also critique “Jurassic World Rebirth,” the latest offering in a multibillion-dollar movie franchise that was born back in 1993.

Holtz, who specializes in the study of meat-eating dinosaurs like the ones that grab the spotlight in the movie, is a co-author of the educational field guides for the early “Jurassic Park” movies as well as the “Jurassic World” sequels that followed. He’s also been a scientific consultant for dino-documentaries including the BBC’s “Walking With Dinosaurs” series and the Discovery Channel’s “When Dinosaurs Roamed America.”

He argues that the Jurassic movies elevated the general public’s long-running fascination with dinosaurs to a whole new level.

“There was a time when no one but paleontologists had ever heard the word ‘Velociraptor,’ and that time was 1993,” Holtz says. “Prior to that, the general public would have never heard of this particular sort of dinosaur, which is now by far one of the most common.”

“Jurassic Park” reflected the state of the art for dinosaur science when it came out, and the special effects have only gotten better with each sequel. But Holtz admits that the movies’ depictions of dinosaurs haven’t always kept pace with what paleontologists have discovered.

“I’ve talked with special-effects people who’ve been involved with some of these movies, and they said, yes, they would like to run a little more wild, a little more modern,” Holtz says. “But there is the pressure to have the look and feel of a Jurassic World / Jurassic Park franchise dinosaur. And you know, that’s kind of a shame.”

There are several ways to go wild with dinosaurs. For example, fossil specimens that were unearthed in China’s Liaoning Province in the 1990s confirmed that non-avian dinosaurs had feathers. And when scientists studied fossilized traces of microscopic structures known as melanosomes, they determined that dinosaurs came in different colors. Such traits aren’t seen so much in the creatures of “Jurassic World.”

“They did start to add patches of colors in more of the dinosaurs in the ‘Jurassic World’ series,” Holtz says. “The raptors — although they’re not feathered as they should be, they at least have more interesting colors on them. In fact, Blue gets its name because of the blue streaks down its body.”

Myhrvold wishes the filmmakers had included some of the more bizarre species that paleontologists have turned up in the fossil record. One such species is Microraptor gui — a type of raptor that was discovered in the early 2000s and had four wings with black feathers.

“There are insects with four wings, like dragonflies, but there aren’t any birds,” Myhrvold says. “Deinocheirus … that’s another raptor relative that everyone thought was going to be the most vicious-looking dinosaur on earth because they found the claws and the arms in the 1970s. They finally found the rest of it, and it turns out it has a face like a horse. There are things like that, that would be cool to see more of … but you know, the goal of the people making the movie is to be entertaining, not to satisfy all the whims of us paleontologists.”

Some of the dinosaur depictions in the latest movie are debatable. For example, one scene has a school of spinosaurs with colorful sails menacing a ship and its occupants on the open sea. Both Myhrvold and Holtz have played a part in research arguing that spinosaurs couldn’t possibly behave that way.

“I believe, and I know Tom believes, that Spinosaurus was a shoreline animal,” Myhrvold says. “It undoubtedly ate fish, but you can eat fish without swimming 43 kilometers offshore. Grizzly bears eat fish, and they don’t swim offshore if they can help it. Now, admittedly, that’s controversial.”

Another slight misstep has to do with titanosaur tails. One scene in “Jurassic World Rebirth” shows a flock of titanosaurs flicking long, sinuous tails as they graze. Myhrvold actually conducted research into dino-tails, focusing on whether some dinosaurs could whip their tails at supersonic speeds. To show that they could, he went so far as to have a mechanical tail built that could simulate the whipping action.

Unfortunately, titanosaurs didn’t have the type of tail that’s shown in the movie. “When I saw the long tail, I first thought, hey, this can’t be a titanosaur,” Myhrvold says. “And then I thought they’re going to crack it supersonically, and I would forgive everything else.”

Myhrvold argues that it’s worth the effort for science-fiction movies to stay as close to the science as possible, with limited exceptions for entertainment value.

“There are movies that really do pride themselves on getting details right — details of spycraft, details of weapons, details of historical facts — and then they spend their suspension-of-disbelief items on a smaller set of things,” he says. “They’re rewarded in having those details be accurate with a lot of people loving that.”

The basic premise of the “Jurassic Park” and “Jurassic World” movies is broad enough to give the filmmakers plenty of literary license. If you’re able to reconstruct dinosaur species from bits of DNA, you should be able to mix and match that genetic code to create new hybrid species.

In the “Rebirth” movie, Scarlett Johansson and a team that includes a greedy business executive and a virtuous paleontologist have to grab blood samples from genetically altered dinosaurs that were too dangerous to be set loose, but too expensive to kill. “The worst of the worst were left here,” says the executive (played by Rupert Friend).

The idea of resurrecting extinct species from their DNA isn’t nearly as far-fetched as it sounded back in 1993: Today, commercial ventures including Colossal Biosciences are working on ways to create modern-day versions of dire wolves, woolly mammoths and even woolly mice. But those efforts pale in comparison with the over-the-top creatures created by the filmmakers behind the “Jurassic World” movies.

Two made-up species take the spotlight in “Rebirth”: Mutadon, which is a cross between a velociraptor and a pterosaur; and Distortus rex, a monster with six arms and an oversized, misshapen head.

Myhrvold thinks it’s unnecessary to make up imaginary monsters. “The part that is funny to me about this is that there are lots more dinosaurs,” Myhrvold says. “It turns out, in most cases, the actual fossil record of Earth is more interesting than made-up creatures.”

The real-world technology of paleontology isn’t quite as advanced as the fictional technology of “Jurassic World,” but it’s come a long way since 1993. Holtz notes that CT scans of dinosaur skulls “have revealed lots of information about their brain cavities and nerve passages … which were extremely difficult to try to see before.”

“Isotopic studies of the bones and teeth can tell us something about what they were feeding on,” Holtz says. “Very clever geochemists and biogeochemists have been able to tease out organic goo from fossils in a way that tells us something about certain aspects of their physiology.”

Even artificial intelligence comes into play. Holtz says some paleontologists are using AI tools to classify their fossil specimens. “Machine learning might be able to pick up things that we don’t immediately think of,” he says. “So far, that’s been the main use of it I’ve seen professionally, but it’s still early days. We’ll see where this technology goes in the future.”

How much of a future is there for paleontology — and for the “Jurassic World” franchise? Will there come a time when dinosaur movies go extinct? During the opening scenes of “Jurassic World Rebirth,” we learn that the general public has become so bored with dinosaurs that the museums are shutting down their exhibits.

“Sorry, I find that deeply unrealistic,” Myhrvold says. “That was maybe the most unrealistic thing in the film. You can give me a six-legged dinosaur, OK? Don’t tell me no one’s gonna like dinosaurs, because dinosaurs have the great additional feature that they’re real. And at some level, the fact that they’re real keeps bringing them home again.”

Dino bites

Mini-review: Setting aside their scientific qualms, Holtz and Myhrvold give two thumbs up to “Jurassic World Rebirth” for its entertainment value. “This is not a film that’s supposed to be deeply philosophical. This is a film that’s supposed to show you lots of action, lots of thrilling sequences, and dinosaurs eating people. And as long as you check each of those, hey, you’re there,” Myhrvold says. Holtz adds that the movie has another key element for dinosaur movies: children at risk. Myhrvold agrees: “Not only do dinosaurs have to eat people … they have to almost eat children, right? If they did eat children, I think that would be off-putting.”

The next ‘Jurassic’ movie? Deadline reports that “Jurassic World Rebirth” has gotten off to a “stomping good start,” and it seems as if the path is open for yet another sequel. “I bet you could build a story around the dinosaur field biologist who’s not working with fossils or bones in plaster jackets — they’re working with live animals, but trying to figure out cool things about them,” Myhrvold says.

Delve deeper into dinosaurs: Check out Myhrvold’s detailed analysis of what “Jurassic World Rebirth” got right and wrong about paleontology, which also includes a lineup card for the movie’s creatures. Holtz provides his own perspective on facts vs. film fiction in a report from the University of Maryland. And for the full treatment, you can turn to the Jurassic World Dinosaur Field Guide, which Holtz had a hand in writing; or “The Rise and Fall of the Dinosaurs,” which was written by paleontologist Steve Brusatte, a science consultant for “Jurassic World Rebirth.”

Fiction Science is included in FeedSpot’s 100 Best Sci-Fi Podcasts. Stay tuned for future episodes of the Fiction Science podcast via Apple, Spotify, Player.fm, Pocket Casts and Podchaser. If you like Fiction Science, please rate the podcast and subscribe to get alerts for future episodes.

Scientists say they’ve put together a new kind of molecular toolkit that could eventually be used to treat a variety of brain diseases, possibly including epilepsy, sleep disorders and Huntington’s disease.

The kit currently contains more than 1,000 tools of a type known as enhancer AAV vectors, with AAV standing for “adeno-associated virus.” A consortium that included researchers from Seattle’s Allen Institute for Brain Science and the University of Washington combined harmless adeno-associated viruses with snippets of engineered DNA to create a gene-therapy package that could target specific neurons in the brain while having no effect on other cells.

Researchers laid out their findings in a set of eight studies published today in the Cell Press family of journals. The work is part of a project called the Armamentarium for Precision Brain Cell Access, funded through the National Institutes of Health’s BRAIN Initiative.

“Honing in on the right cells — in the right way and at the right time — is the future of precision brain medicine,” John Ngai, director of the BRAIN Initiative, said in a news release. “These tools move us closer to that future, while also expanding what we know about the brain’s cells and circuits today.”

Jonathan Ting, one of the study authors and an associate investigator at the Allen Institute, noted that enhancer AAV vectors have been used before to target cells in the brain and other tissues such as the heart and liver. In March, for example, a team led by researchers at the Allen Institute and Seattle Children’s Research Institute described an AAV vector that shows promise for treating Dravet syndrome, a genetic disorder that causes a severe form of epilepsy.

“What is unique about this new collection is the immense scale of new cell-type targeting tools,” Ting told GeekWire in an email. “Rather than a trickle of tools for the purpose of advancing understanding of the brain, now we have what amounts to a tidal wave of new tools to access and perturb cell function across a remarkable diversity of cell types in multiple brain regions and the spinal cord.”

Different enhancer AAV vectors were engineered to target different cell types in different regions of the central nervous system, including the spinal cord, the cortex and the striatum. That specificity could open new avenues for targeted treatment. For example, dysfunction in the striatum is linked to movement disorders such as Parkinson’s and Huntington’s disease — and may also play a role in drug addiction.

“There is an overall principle that diseases usually arise from flaws in specific cell types, not the whole organism. For example, epilepsy is a nervous system disease that is actually a disease of specific neurons in the nervous system,” said study author Bosiljka Tasic, director of molecular genetics at the Allen Institute. “If you want to fix those neurons, you can try to access only those neurons. The key is this cell-type-specific access for understanding and perturbing brain cells to figure out their function, and for correcting and fixing the defective parts of these cells.”

One study documented how one of the molecular tools targeted a rare type of cell that regulates sleep. “It is reasonable to speculate that this tool and various derivative tools could be used to better understand the central control of sleep, or to intervene in cases of sleep disturbance disorders in humans,” Ting told GeekWire. “The exact applications remain to be determined.”

The experiments described in the studies were conducted on mice, rats and macaque monkeys, plus tissue samples from those species as well as from marmosets and humans. One of the studies used machine-learning models to identify the most promising enhancer for targeting a specific region of the rhesus monkey brain. That region, known as the dorsolateral prefrontal cortex, is associated in humans with working memory and impulse control.

In the short term, the new toolkit should help researchers use animal models to gain new insights into the workings of the different types of cells in the central nervous system, and learn how to switch specific functions of those cells on or off. Gordon Fishell, a professor of neurobiology at Harvard and the Broad Institute, said that gaining access to a variety of cell types will be a “game-changer in understanding the brain and developing therapies for human neurological disorders.”

It may take a while for clinical applications to emerge. “The time horizon is unclear for how fast any of these novel approaches can advance to the clinic, but we expect to see advances along these lines relatively quickly, as in the next two to three years,” Ting said. “This may take the form of developing a ‘Version 2.0’ approach to a failed Version 1.0 strategy that didn’t target cell types and yielded some disappointing outcomes.”

Ting and his colleagues expect to improve the safety and efficacy of their tools as they become more familiar with the toolkit. “We expect this scenario will play out for many diverse CNS [central nervous system] disorders into the future,” he said.

In addition to the Allen Institute for Brain Science and the University of Washington, the institutions collaborating in the studies published today include the Broad Institute, Harvard Medical School, Duke University, the University of California at Irvine, the University of California at Berkeley, the University of Pittsburgh, Carnegie Mellon University, Stanford University and Addgene. The tools and data in the studies are freely available on the Allen Institute’s Genetic Tools Atlas and through Addgene.

Here are the papers published today related to the BRAIN Armamentarium:

• Neuron: “Exploring brain circuits, one cell type — or more — at a time”
• Cell: “A suite of enhancer AAVs and transgenic mouse lines for genetic access to cortical cell types.”
• Cell Genomics: “Evaluating methods for the prediction of cell-type-specific enhancers in the mammalian cortex.”
• Cell Reports: “Enhancer AAVs for targeting spinal motor neurons and descending motor pathways in rodents and macaque.”
• Cell Reports Methods: “An AAV capsid proposed as microglia-targeting directs genetic expression in forebrain excitatory neurons.”
• Neuron: “An enhancer-AAV toolbox to target and manipulate distinct interneuron subtypes.”
• Neuron: “Enhancer AAV toolbox for accessing and perturbing striatal cell types and circuits.”
• Neuron: “Specific targeting of brain endothelial cells using enhancer AAV vectors.”
• Neuron: “Machine learning identification of enhancers in the rhesus macaque genome.”

Seven years after they started, neuroscientists have published the results of a landmark study that was designed to determine which theory of human consciousness came closest to the mark — and those results are decidedly mixed.

The bad news is that neither of the leading theories held a clear advantage in explaining how consciousness arises. The good news is that researchers picked up new clues about where to look.

One of the leaders of the effort — Christof Koch, a meritorious investigator at the Seattle-based Allen Institute — said he was heartened by the state of the debate.

“Adversarial collaboration fits within the Allen Institute’s mission of team science, open science and big science, in service of one of the biggest, and most long-standing, intellectual challenges of humanity: the Mind-Body Problem,” Koch said in a news release. “Unraveling this mystery is the passion of my entire life.”

The history of the study that was published today in the journal Nature goes back to a workshop that took place at the Allen Institute in 2018. The discussion focused on two different hypotheses about the nature of consciousness.

One hypothesis, known as Integrated Information Theory or IIT, holds that consciousness arises through the interaction and cooperation of various parts of the brain as information is processed.

“IIT suggests that consciousness emerges when information inside a system (like the brain) is highly connected and unified; and as long as the information remains unified, it will be consciously perceived,” Koch, who has been a proponent of IIT, told GeekWire in an email. “IIT starts with consciousness and then infers the causal property any mechanism, such as a brain, needs to obey to be conscious.”

The other leading hypothesis is called Global Neuronal Workspace Theory. “GNWT suggests a network of brain areas spotlight important pieces of information in the brain — bringing it to the forefront of our minds — broadcasting it widely the moment it enters consciousness, and this produces conscious experience,” Koch said. “GNWT starts with psychology and neuroscience, and seeks to distill consciousness from these.”

Was it possible to find evidence favoring one hypothesis or the other? In 2019, Koch and other researchers worked out a series of tests that they thought could address the issue. The Cogitate Consortium recruited 256 human subjects across 12 laboratories to have their brain activity recorded.

The research team monitored the brain’s blood flow, electrical activity and magnetic resonance readings while the experimental subjects interacted with a series of on-screen images of faces, objects, letters and symbols. All those data points were analyzed to track how impulses moved between regions of the brain.

If the neural nexus of consciousness could be traced to the prefrontal cortex, that would favor the GNWT hypothesis. But if the brain activity associated with consciousness was centered toward the back of the brain, in the posterior cortex, that would argue against GNWT — and in favor of IIT.

“Here the evidence is decidedly in favor of the posterior cortex,” Koch said. “Either information pertaining to the conscious experience couldn’t be found in the front, or it was far weaker than in the back. This supports the idea that while the frontal lobes are critical to intelligence, judgment, reasoning, etc., they are not critically involved in seeing, in conscious visual perception.”

The results of the experiment also suggested the brain activity that’s correlated with consciousness continued as long as a conscious experience was sustained. That was another strike against GNWT, which predicts that such activity should peak at the beginning and end of an experience, but fall off in the middle.

In a third experimental faceoff, however, GNWT had the upper hand: The researchers found that there was a high-frequency oscillation between the brain’s early visual cortical regions and the front of the brain — but not between those early visual regions and the back of the brain, as predicted by the IIT hypothesis.

The Cogitate Consortium provided a progress report on the experiment in 2023, acknowledging that there wasn’t a clear winner but pledging to continue the study. That lack of a clear result caused Koch to lose a decades-old bet with a philosopher — a wager that he paid off with a case of fine wine.

In this week’s Nature research paper, the consortium’s scientists acknowledged that they still haven’t settled on a winner. Nevertheless, they gleaned useful information about the regions of the brain that appear to be associated with consciousness.

“We found information about conscious content in visual, ventrotemporal and inferior frontal cortex, with sustained responses in occipital and lateral temporal cortex reflecting stimulus duration, and content-specific synchronization between frontal and early visual areas,” the researchers reported.

Koch said there are plenty of reasons for continuing to delve into one of life’s deepest mysteries.

“If we want to understand the substrate of consciousness, who has it (adults, pre-linguistic children, a second trimester fetus, a dog, a mouse, a squid, a raven, a fly, etc.) we need to identify the underlying mechanisms in the brain, both for conceptual reasons (why these and not other ones) as well as for clinical ones,” he said in his email.

Being able to determine the state of mind of an unresponsive patient could make a critical difference when it comes to deciding whether or not to withdraw life-sustaining treatment. Some studies have suggested that more than half of the deaths that take place in critical care occur following withdrawal of treatment. But Koch cited a study reporting that around a quarter of unresponsive patients are actually conscious but “unable to signal this at the bedside.”

Koch is one of the founders of Intrinsic Powers, a venture that is working on a medical device that could serve as a “consciousness-meter” for unresponsive patients. The device uses trans-cranial magnetic stimulation to stimulate the brain, and then checks for a response. “Whether or not IIT is the correct theory of consciousness, the device works as intended,” Koch said.

Even though the results of the years-long experiment in adversarial collaboration turned out to be inconclusive, Koch hasn’t given up on IIT. “The challenge is to come up with ways to explain the null result in the third prediction,” he said. “We’re working on that.”

Koch is among more than 40 authors of the Cogitate Consortium study in Nature, titled “Adversarial Testing of Global Neuronal Workspace and Integrated Information Theories of Consciousness.” The study’s corresponding author is Lucia Melloni of the New York University Grossman School of Medicine.

University of Washington biochemist David Baker is a hot ticket. Baker, who won the Nobel Prize in chemistry in October, straddles the worlds of academia and entrepreneurship. He leads the UW Medicine’s groundbreaking Institute for Protein Design and has co-founded 21 biotech companies.

On Wednesday, he gave the keynote presentation to a packed audience at the Life Science Innovation Northwest conference — and dashed out at the end to catch a flight for dinner with former President Barack Obama.

Like we said, hot ticket.

At the Seattle conference, Baker addressed the future of using artificial intelligence to create previously nonexistent proteins with the potential to provide new health care treatments and address environmental damages. He also spoke of challenges, including his deep concerns about the impacts of the Trump administration’s funding cuts on the next generation of scientists.

Read on for other highlights from the conversation, which was moderated by Simone Fishburn, vice president and editor-in-chief of the analytics company BioCentury.

On the evolution of protein design

Baker recalled a time when creating de novo proteins as therapeutics was essentially viewed as “the lunatic fringe.”

Back around 2010, his team pivoted from the idea of “taking naturally occurring proteins and trying to redesign them to do other things. [Instead] we were going to learn how to build proteins completely from scratch,” he said. “I think everyone laughed like, ‘that’s crazy.’

“But now, every day I read something about a new protein design company starting — people I’ve never met — and they always talk about how they’re joining the protein design revolution,” Baker said.

Why this innovation is based in academia

While some expected industry and efforts like Google’s DeepMind to dominate protein design, the innovation is coming from academia.

“Protein design is, how do you make new proteins?” Baker said. “Well, there are a couple of problems there. First of all, you have to actually make them in the lab, which means you have to have a lab, to know how to do an experiment. Second, you have to know what to make. And so you have to have some understanding of biology.”

University researchers in particular had ideas about what problems to go after.

“That high level thinking was very important,” Baker said.

Where is the protein design field headed

Baker highlighted two areas for development. The first is tweaking already understood proteins to optimize their performance for specific health applications.

The second is generating new kinds of therapeutics that perform multiple, complicated functions. That includes proteins that recognize, bind and destroy a target, as well as proteins that act like “machines.” These therapeutics would use other molecules in circulation as tools to help it address an ailment.

“That’s a little bit more futuristic,” Baker said.

On Seattle as a protein design hub

Baker, who grew up in Seattle and was eager to return to the city, recalled frustrating conversations with investment firms that took place in past decades. “We never got anywhere,” he said, “because I always insisted the company be in Seattle. And they said, ‘Oh, we can’t start companies in Seattle.’

“So now it’s completely different,” Baker continued. “When I talk to a VC firm about starting a company, they say, ‘Well, where in Seattle?'”

Now the city “has a higher concentration of really world-class expertise in protein design, in terms of workforce, than anywhere else,” he said. “And so that’s great for companies starting off. The companies also can help each other in many ways.”

The joy of a clouded crystal ball

“What I’ve always said is that I can’t predict the future of science,” Baker said. “I’m very glad I can’t, because it’d be totally boring if I could … I mean, my goal is that at any given point, I just can’t even imagine what we’ll be doing two years later. And I think that’s held up pretty regularly.”

At the risk of sounding terribly reductionist, one could describe the Washington State Academy of Sciences as a super precise AI agent, ready to provide lawmakers and state officials with scientific insights on pressing public issues.

More accurately it’s a nonpartisan, publicly funded nonprofit organization with more than 400 highly distinguished members who are the experts in their fields. Participants are nominated by peers to share their wisdom with the Legislature, the governor’s office and state departments. The science academy holds individual and small meetings, roundtables and symposia. It prepares technical reviews and reports.

And it recently appointed Melanie Roberts as its executive director.

Roberts holds a doctoral degree in neurobiology and behavior from the University of Washington, was a director of state and regional affairs at the Pacific Northwest National Laboratory, and created an emerging leader program at the American Association for the Advancement of Science, among other roles.

“I can’t imagine any place that would be a better fit with what I’ve been trying to do for a very long time,” she said, “which is better connecting science with societal outcomes.”

In one recent example, the group received an inquiry from state leaders about specific toxic ingredients in a consumer product. The academy quickly convened a trio of researchers, one in human toxicology, another in manufactured materials and a third in regulatory science.

“That’s one of the things that I think makes us different,” Roberts said. “We can look at issues from lots of different angles.”

GeekWire recently caught up with Roberts to learn more about the 20-year-old organization during a time when innovation is steaming ahead, but U.S. officials are cutting research funding and confidence in scientists has declined — though three-fourths of Americans still trust them.

Here are highlights from our conversation, edited for clarity and length.

GeekWire: How does WSAS operate?

Roberts: In the past, we’ve done the studies, we’ve done workshops, which include state agency folks and governor and tribal representatives, and then we’ve done our annual symposia. We have done some public events, but mostly we focus on education for the state [officials].

If there’s a piece of legislation that is being deliberated, for example, do you get tapped to provide expertise?

One of the things we’re hearing, especially from the Legislature, is: “We need a rapid response. People are sending us studies and making claims that we don’t know how to evaluate, and can you help us?” So this is a place where we’re looking to expand our offerings, this rapid response service.

Let’s talk about what your organization is doing in a time when research funding is being cut.

Certainly there are going to be gaps in federal funding in a number of areas, not just science. We are a network of science and technology experts, so we’re thinking, how can we fill gaps? For example, if there are certain studies or efforts that had been led at the federal level in the past, and Washington then doesn’t have the information that it needs, what can we do to help fill those gaps?

Computer science and technology are moving at breakneck speed. Are there areas where you’re anticipating or already seeing a lot of demand for your experts?

Our symposium last year was on artificial intelligence, because obviously that’s an area that everybody is thinking about right now. Other areas we hear about are quantum computing. Energy is a big one for this state and all of the different energy technologies — [questions like] when is fusion going to come online?

This is a challenging time for the scientific sector in the U.S. Are there things that make you hopeful?

There’s always an opportunity for innovation during times of change. And so how does Washington do that? How do we, as a state, think about moving into this new era? We’ve done it many times before. There’s the Gold Rush and there’s the aviation innovation boom. We’ve gone through a lot of eras, and so how can Washington survive, and even thrive, in this new era? We’ve got such an innovative spirit. We’ve got a great research community. We’ve got really creative people. So in hardship, there are opportunities.

Researchers say they’ve accomplished a feat that was said to be impossible 46 years ago: mapping the cells in a cubic millimeter of brain tissue and tracing their activity.

The achievement, documented today in a set of research papers published by the Nature family of journals, is being compared to the Apollo moon shots that were launched more than 50 years ago, and to the drafts of the human genome that were released more than 20 years ago.

Scientists from Seattle’s Allen Institute played a key role in the $100 million effort known as the Machine Intelligence from Cortical Networks program, or MICrONS. More than 150 researchers worked together through MICrONS to create a detailed 3D map of a cubic millimeter taken from a mouse’s brain — and figure out how the 200,000 brain cells in a speck the size of a coarse grain of sand work together.

“It really has been one of the holy grails of the field from the beginning,” Clay Reid, a senior investigator at the Allen Institute, told GeekWire. “There are many thousands of neuroscientists who study the cerebral cortex, and pretty much everyone who studies the cerebral cortex would like to be able to know what are the sources of inputs to any given cell within the cortex, and what are the outputs of that cell. That’s what such a complete data set allows one to study.”

The origin story for this particular holy grail goes back to 1979, when Francis Crick, the co-discoverer of DNA’s double-helix structure, mused about the promise of neuroscience — and about the field’s limitations. “It is no use asking for the impossible, such as, say, the exact wiring diagram for a cubic millimeter of brain tissue and the way all its neurons are firing,” Crick wrote in Scientific American.

That challenge struck a chord with Reid, who was a college student at the time. He set out to prove Crick wrong, and succeeded. “That’s exactly the experiment that we just finished up,” Reid said.

The project’s first steps were taken at Baylor College of Medicine in Texas, where scientists used specialized microscopes to record the brain activity from a cubic millimeter’s worth of a mouse’s visual cortex as the animal watched movies and YouTube clips. That bit of tissue was then sent to the Allen Institute, where it was sliced into more than 25,000 thin layers. About 95 million high-resolution images of the tissue slices were recorded using an array of electron microscopes. Finally, researchers at Princeton University used artificial intelligence tools to turn the images into a 3D reconstruction of the tissue sample on a cell-by-cell basis.

The wiring diagram and its supporting files amount to 1.6 petabytes’ worth of data. The map traces more than 2.5 miles (4 kilometers) of tangled-up axons, the fibers that serve as the “wiring” for brain cells. It pinpoints 523 million synapses, which are the connection points between cells. Just as importantly, the map provides a guide to the activity patterns recorded by the Baylor team.

Over the years, MICrON’s researchers have provided progress reports on the project, but the studies published today in Nature and its sister journals serve to sum up their work. The papers present findings about the structure of the visual cortex, the assortment of cells found in the sample and how those cells function.

One of the more significant findings has to do with how inhibitory cells control the activity of other cells in a neural circuit. “They’re certainly not on-off switches for the entire circuit,” Reid said. “Different types of inhibitory neurons inhibit different elements within the circuit. They’re switches, but they’re very carefully wired. They don’t turn on and off every light in the building.”

In the years ahead, neuroscientists could use the freely available MICrONS data set to fine-tune their models of brain structure and function. It might also be possible to track down the causes of, and potential treatments for, neurological conditions ranging from Alzheimer’s and Parkinson’s disease to schizophrenia and autism.

“If you have a broken radio and you have the circuit diagram, you’ll be in a better position to fix it,” MICrONS team member Nuno da Costa, an associate investigator at the Allen Institute, said in a news release. “We are describing a kind of Google map or blueprint of this grain of sand. In the future, we can use this to compare the brain wiring in a healthy mouse to the brain wiring in a model of disease.”

The data set could also point the way to innovations in artificial intelligence — perhaps including a new generation of neuromorphic computers that would process data the way biological brains do.

Reid pointed out that MICrONS was funded by the federal government through the BRAIN Initiative and the Intelligence Advance Research Projects Activity, or IARPA, partly to seek out new strategies for AI. “The goal of this was, why don’t we use the most complete and detailed characterization of a cortical circuit perhaps as inspiration for new architectures for machine learning?” he said.

David Markowitz, the former IARPA program manager who coordinated the MICrONS program, characterized the funding as a “moonshot investment.” He said the research papers published today mark “a watershed moment for neuroscience, comparable to the Human Genome Project in their transformative potential.”

As was the case for the Human Genome Project, it will take years for the MICrONS data set to settle into its final form. “Yes, we have the morphology for all of the neurons,” Reid said. “Yes, machine learning has located and identified all of the synapses. But the final step of having humans verify every connection in that wiring diagram has not been done. … In order to get there, we will need advances in machine learning.”

The National Institutes of Health is already looking ahead to future frontiers in brain mapping with a program called BRAIN CONNECTS. (That’s a tortured acronym that stands for BRAIN Initiative Connectivity Across Scales.) Two of the goals of that program are to generate a detailed wiring diagram for the complete mouse brain, and to map long-distance connections between different areas of the human brain.

So, what about mapping the entire human brain? “Because of the size of the human brain, it is unimaginable, and I would say impossible in any reasonable future, to map the entire human brain at the level that one did for this cubic millimeter for the MICrONS project,” Reid said.

When he was reminded that Francis Crick said the same thing about mapping that cubic millimeter back in 1979, Reid expanded upon his remarks.

“Crick never set an expiration date for his pronouncement,” he said. “It is possible that we could do this for the human brain, but from this viewpoint, it still is unimaginable. A lot can happen in 46 years. Certainly a lot has happened in the 46 years since Crick said that something was impossible.”

Nature has set up a landing page for research papers related to the MICrONS project. Other papers are being published in Nature Communications and Nature Neuroscience. The MICrONS consortium includes scientists and researchers from the Allen Institute, Baylor College of Medicine, Princeton and many other institutions. The Allen Institute has created a multimedia presentation that traces the history of the project.

Since 2014, UW Medicine’s Institute for Protein Design has spun off 10 startups, and the institute’s director, University of Washington biochemist David Baker, has co-founded 21 tech companies. After Baker won the Nobel Prize in October, interest spiked from graduate students and postdoctoral researchers eager to join IPD in pursuit of academic or entrepreneurial ventures.

The timing was perfect. The artificial intelligence used to create previously nonexistent proteins with the potential to provide new cancer treatments, tackle plastic waste, deliver life-saving vaccines, capture carbon and other essential tasks has hit its stride.

Then everything came to a screeching halt.

Due to Trump administration efforts to slash research budgets and threats to state funding, UW Medicine in February instituted a temporary hiring freeze for non-clinical roles, followed by a university-wide freeze on non-essential positions in March.

That means UW academic departments will likely accept fewer graduate students, and labs — Baker’s included — are unable to hire postdocs unless they successfully lobby leadership for exemptions. That means fewer highly educated researchers will get the chance to delve into these new technologies, potentially curbing the startups spinning out of IPD.

“In some sense it’s like [Charles] Dickens wrote — ‘It was the best of times, it was the worst of times,'” Baker said in an interview with GeekWire.

“The opportunities are incredible. The amount of talent that would like to come here to work on these problems is enormous,” he said, thanks in part to the Nobel Prize. But with the freeze and other funding cuts, they’ve had to push applicants away.

‘A disaster’

If the timing had been different, David Younger could have been one of those people.

Younger is the CEO of A-Alpha Bio, one of the startups that launched out of IPD research. He did his PhD and postdoc at the UW, and co-founded Seattle’s A-Alpha Bio in 2018 after incubating at the UW’s CoMotion. The company is designing proteins and measuring their interactions with other proteins, looking for new connections such as those between antibodies and viruses.

The company has raised $65.5 million, including venture capital investments, Department of Defense funding and grants. It employs about 50 people.

Younger is troubled by diminished support for academic training and research, and its broader impacts.

“It’s a disaster for everyone,” he said. “There are so many different types of careers that are enabled only by those types of educational opportunities and experiences.”

The Trump administration has chopped away at university budgets through multiple avenues. In the name of efficiency, it wants dramatic cuts to National Institutes of Health reimbursements for costs including rent, utilities, administrative support and other infrastructure. Research funding at Columbia University and Harvard is facing uncertainty, with huge amounts of federal support withdrawn or under review amid allegations of antisemitic discrimination. The University of Washington is among the institutions the Trump administration may also penalize.

Added to that uncertainty is a revenue shortfall in Washington state that is pushing Gov. Bob Ferguson and state lawmakers to shrink spending.

“The cuts proposed by the Governor and state legislature disproportionately impact the UW — and do so at a time when the federal government’s damaging cutbacks to research and health-care funding are already hitting home, magnifying their impact,” UW President Ana Mari Cauce wrote in an email to alumni on Tuesday.

Potential for lost leadership

Marc Cummings, CEO of Life Science Washington, an organization promoting the state’s biotech sector, fears the economic impacts of the cuts. In this state, the industry “is primarily home grown with most local companies being able to trace their success back to technology or people that came from UW or other local research institutes,” he said by email.

“America has built a life science industry that is the envy of the world,” he added. “The industry employs 2.3 million people in the U.S. and we lead the world in biotech investment and new drug development. Why? Because we have the best research universities that attract the best global talent and conduct the most advanced medical research.” 

Helen Eisenach is a grad student working with Baker and IPD associate professor Neil King. She has funding for about another year-and-a-half to work on her PhD. What she does next is increasingly uncertain given the decreasing federal support for research.

The situation puzzles her.

While Eisenach is focused on engineering new proteins for treating health and environmental issues, science also stokes progress in communications, technology, military advancement, agriculture and manufacturing. Without funding, the U.S. gives up its leadership in these fields, she said.

“You’re up against countries,” Eisenach said, “who have absolutely no qualms about spending every dollar towards being the best in class.”

Making the world a better place

Daniel Adriano Silva co-founded two companies with IPD roots: Neoleukin Therapeutics, which launched in 2019 and later merged with Neurogene, and Monod Bio, a 20-person company started in 2021 that creates proteins for use in research, health diagnostics and clinical settings.

Silva participated in IPD’s Translational Investigator program, which helps academics transition into entrepreneurship.

He is now CEO of Seattle-based Monod, and believes the local biotech ecosystem is special.

“There is a concentration of talent around machine learning and AI and biosciences here that is unique,” Silva said.

Labs across UW Medicine, the UW and elsewhere in the U.S. are waiting to see how the funding questions are settled and how many researchers they’ll be able to support. Baker weighs the pros and cons of the situation he faces.

“It’s a big blow, just because everything was so poised and the opportunities are so great,” Baker said. With the uncertainty, he wonders “can we support the people who are here? Can we bring in new people? Can we really capitalize on all these amazing possibilities for designing proteins to make the world a better place?

The Dickens quote cited by Baker — which is the start of “A Tale of Two Cities” — continues: “it was the age of wisdom, it was the age of foolishness, it was the epoch of belief, it was the epoch of incredulity, it was the season of light, it was the season of darkness, it was the spring of hope, it was the winter of despair.”

RELATED: For Nobel Prize winner David Baker, it’s back to the lab as protein pioneer helps launch more startups

A team from the University of Washington led by Nobel laureate David Baker is using artificial intelligence to design effective enzymes from scratch — an accomplishment the researchers call “one of science’s grand challenges.”

Enzymes are the wizards of the natural world, proteins that can transform molecules and rapidly accelerate chemical reactions under mild conditions. They’re found in every living cell and are essential to life. Enzymes are already being harnessed for drug production and industrial processes. The newly developed tools for enzyme creation could unlock wide-ranging applications.

“Now we can make these enzymes tailored to any reaction of interest, theoretically,” said Anna Lauko, a recent Ph.D. graduate from Baker’s lab. “It’s sort of flipped the way that we would do enzyme design.”

Lauko is a co-lead author of a research paper being published today in the journal Science. Her co-leads are Sam Pellock, an acting instructor in the lab, and Kiera Sumida, one of Baker’s graduate students.

Last year Baker, a biochemist and director of the Institute for Protein Design at UW Medicine, won a Nobel Prize in chemistry for his work unraveling the molecular design of proteins and developing technologies for using AI to build and test new ones.

In the past, scientists made Frankenstein enzymes, stitching together components of existing proteins in the hope that the assembled parts could manage a precise task. But enzymes often need to perform nuanced operations, changing shapes multiple times as they manipulate molecules.

Pellock compared the old approach to enzyme design to going to a thrift store for a suit.

“It’s unlikely that you’re going to find a suit that fits well,” he said, and the enzymes were the same way. They included the basic pieces, but didn’t perfectly match the molecules that they needed to interact with. The new approach produces bespoke proteins.

To test their cutting-edge approach, the UW researchers focused on a well-studied enzyme called a serine hydrolase. The enzyme is able to cleave a chemical bond that’s key to the structure of many carbon-containing molecules, including plastics, polyesters and a common fat in humans.

The team used the RFdiffusion model, an AI program for generating proteins that was previously developed by Baker’s lab and is open source. They combined that with a newer tool called PLACER that helped them identify the most promising de novo enzyme candidates. The scientists then tested the performance of the machine-created enzymes.

“They’re still not quite as good as native enzymes,” Pellock said. “But out of the computer, these are among the best that have been made and they were made with very high accuracy.”

The accomplishment is a milestone and proves that the researchers are getting close to making new enzymes for human-driven tasks that could outperform what nature has produced.

Sumida, for example, is working to build an enzyme that could help degrade the planet’s massive glut of plastic waste. Plastic is an incredibly new substance on an evolutionary scale so there hasn’t been much time for enzymes to evolve that can break it down.

There is an enzyme in the serine hydrolase family that can chop up the bonds in the plastic that’s used to make water bottles and other products, but there are many other kinds of plastics out there that need to be disposed of sustainably.

“We thought that it would be a really good application if we’re able to build these enzymes from scratch,” she said, and customize them for different types of plastic.

The researchers are eager for the advent of high-performing, AI-designed enzymes after decades of largely disappointing efforts.

“Hopefully you’ll start hearing more about enzyme design projects,” Pellock said, “because they’ll actually yield a functional enzyme at the end of them.”

Additional authors for the Science paper are David Baker, Ivan Anishchenko, David Juergens, Woody Ahern, Jihun Jeung, Alex Shida, Andrew Hunt, Indrek Kalvet, Christoffer Norn, Ian Humphreys, Cooper Jamieson, Rohith Krishna, Yakov Kipnis, Alex Kang, Evans Brackenbrough, Asim Bera, Banumathi Sankaran and K. N. Houk.

Author affiliations include the following UW departments and programs: biochemistry; biological physics, structure and design; Institute for Protein Design; molecular engineering; chemistry; the Howard Hughes Medical Institute; and the Paul G. Allen School of Computer Science and Engineering. A scientist from UCLA’s Department of Chemistry and Biochemistry also contributed.

The collaboration is part of NASA’s Transform to Open Science Initiative, which aims to make more than 100 petabytes of satellite imagery and other Earth science data more readily available to scientists, educators, policymakers and the general public.

“The vision behind this collaboration was to leverage AI and cloud technologies to bring Earth’s insights to communities that have been underserved, where access to data can lead to tangible improvements,” Minh Nguyen, Cloud Solution Architect at Microsoft, said today in a blog posting about the project. “By enabling users to interact with the data through simple, plain language queries, we’re helping to democratize access to spaceborne information.”

Tyler Bryson, Microsoft’s corporate vice president for U.S. health and public sector industries, said Earth Copilot lets users interact with NASA’s data repository conversationally.

Bryson noted that the sheer scale and complexity of the information gathered for NASA’s Earth Science Data Systems Program can sometimes make finding the right information a daunting task. Earth Copilot doesn’t require users to navigate highly technical interfaces or master the intricacies of geospatial analysis.

“Instead, they can simply ask questions such as ‘What was the impact of Hurricane Ian in Sanibel Island?’ or ‘How did the COVID-19 pandemic affect air quality in the U.S.?'” Bryson said. “AI will then retrieve relevant datasets, making the process seamless and intuitive.”

The project leverages Microsoft’s Azure cloud platform and the customizable capabilities of Azure OpenAI Service.

“Azure’s robust suite of services, including machine learning, data analytics and scalable cloud infrastructure, powers this AI prototype,” said Juan Carlos López, a former NASA engineer who is now an Azure Specialist at Microsoft. “We’ve designed the system to handle complex queries and large datasets efficiently, ensuring that users can quickly find the information they need without getting bogged down by technical complexities. Our goal was to create a seamless, scalable solution that could evolve as NASA’s data, tools and applications grow.”

Earth Copilot is currently being tested by NASA researchers to explore its capabilities. After the platform is evaluated and fine-tuned, the NASA IMPACT team will look into making it part of its Visualization, Exploration and Data Analysis Project — also known as VEDA.

Some say artificial intelligence will be humanity’s greatest helper. Others warn that AI will become humanity’s most dangerous rival. But maybe there’s a third alternative — with AI agents achieving the status of personhood alongside their human brethren.

The potential for that scenario is the focus of a newly published book titled “The Line: AI and the Future of Personhood.” The author, Duke University law professor James Boyle, says the book has been more than a decade in the making — which suggests more than the usual prescience about the tech world’s current fascination with AI.

In the latest episode of the Fiction Science podcast, he recalls the reaction he received when he shared his early ideas about the book with federal judges more than a dozen years ago..

“They’re like, ‘Rights are reserved for humans, naturally born of women!’ OK, well, not necessarily a great crowd,” says Boyle, founder of Duke Law School’s Center for the Study of the Public Domain. “Obviously, things have changed since then. The book seems perhaps less unhinged now than it did then.”

AI has come a long way since 2011. The smarts that were developed for advanced chatbots like OpenAI’s ChatGPT and Anthropic’s Claude are now being integrated into a wide array of software products, including Microsoft’s Copilot and Amazon’s Alexa voice assistant.

The best AI agents can sometimes sound all-too-human, as a New York Times columnist discovered when a Microsoft chatbot said it was in love with him and wanted to become a human. This year, Stanford researchers declared that ChatGPT-4 has passed the Turing Test — an exercise that was proposed decades ago to define the line between human intelligence and machine intelligence.

Will there come a time when AI’s cognitive capabilities clearly exceed ours — and if so, what implications will that have for the rights and responsibilities accorded to AI agents? Boyle thinks that intelligent machines will eventually achieve some form of personhood.

When he began working on the book, Boyle assumed that AI entities would become so humanlike that they’d have to be granted personhood on moral grounds. An example of that from science fiction would be Data’s effort to win the right of android self-determination in an episode of “Star Trek: The Next Generation.”

“I thought that would be where the debate was,” Boyle says. “And then, increasingly, I realized that an equally likely — maybe even more likely — way that we will actually get some form of legal personality for AI is the same way we did with corporations. Which is, ‘This is convenient. We need someone to sue and be sued.’”

Corporate personhood is a somewhat controversial legal concept that grants corporations some of the rights and responsibilities that flesh-and-blood citizens have (for example, the right to make contracts, and the liability to be prosecuted for crimes) while lacking others (such as the right to marry, run for office or vote).

The idea that Alexa could file suit to get itself out of a contractual obligation may sound as much like science fiction as the Starship Enterprise’s warp drive. But Oren Etzioni, the founder of TrueMedia.org and the former CEO of the Seattle-based Allen Institute for Artificial Intelligence, says it’s not so far-fetched.

“If a corporation is a person of sorts, then certainly we will reach a point where AI agents are,” Etzioni told me in an email. “Once AI agents achieve consciousness, then we will need to treat their feelings with compassion.”

Will ChatGPT and other large language models, or LLMs, ever achieve consciousness? That’s not in the cards, according to Christof Koch, a neuroscientist at Seattle’s Allen Institute (which is separate from the Allen Institute for AI).

“There are two questions really here,” Koch told me in an email. “1. Will the behavior, including language comprehension and speech, of AI/robots become so similar to people’s (or even better in functional terms) that most people will treat them as conscious? 2. Will it ever feel like anything to be such an AI/robot?”

Koch says the answer to the first question is, yes, most likely. “The only uncertainty is by when?” he wrote. “And, as Oren implies, this will have moral, ethical, legal and societal consequences.”

However, Koch argues that the architectures that form the basis for today’s computers are incapable of supporting anything like human-level consciousness. “That is, it will never feel like anything to be a LLM, even though these may achieve, sooner or later, super-intelligence. This may change with radically different hardware, such as quantum or neuromorphic computers,” he said.

“Intelligence is about doing, while consciousness is about being.”

Boyle is sympathetic to Koch’s point of view on the question of whether AI agents will ever become conscious. “The current architecture is ‘no,’ but in the future, I think ‘yes,’” he says.

In his book, and in the podcast, Boyle turns to science fiction as well as science facts to lay out his case. He says the movie “Blade Runner” and the Philip K. Dick novel on which it was based — “Do Androids Dream of Electric Sheep” — serve as “the source of the Nile” for his views on how AI agents of the future might be treated.

Boyle is particularly intrigued by a questionnaire that’s used by detectives in the movie and in the novel to determine whether the entity they’re facing is a natural-born human or an artificially produced replicant. The Voight-Kampff test is meant to measure how much empathy the entity feels in response to emotional stimuli.

“For me, the question that both of those works present is whether we’re all replicants — or rather, whether we can pass any test that we would actually be willing to set for something else,” Boyle says.

Boyle thinks the humans who create AI products, and the consumers who use them, will face a different kind of test in the years ahead.

“I would bet you that sometime within this century, there will be companies — whether it’s Microsoft, or whether it’s Anthropic, or whether it’s OpenAI — who are deliberately saying our self-actualizing AIs, which we treat as people, are participating in this,” he says. “It’ll be kind of like a Whole Foods vibe — a sort of ‘fair-trade coffee’ kind of thing.”

Boyle is also betting that other companies will market their AI agents as “loyal cybernetic servants, and nothing else … you need treat them no more thoughtfully than you do your vacuum cleaner.”

“I would expect that to be a division in the market, like the division between proprietary and free and open-source software,” he says.

When will that moment in the market come about? Which approach will win out? How long until we know whether or not our AI creations fall within the line that marks the boundary of personhood? It’s hard to predict precisely what the timeline might be — or what intelligent machines might do once they cross the line.

“I’ll be very nice to my Roomba in the meantime,” Boyle says. “I think that’s the only thing we can all do.”

“The Line: AI and the Future of Personhood” is available in print and as an e-book. As former board chair of Creative Commons, James Boyle is a proponent of open access to information, and so his book is freely available online via the Duke Law School website. (For what it’s worth, James and I share the Boyle family name but are not closely related.)

Allen Institute neuroscientist Christof Koch has also published several books on the subject of consciousness. His latest book is “Then I Am Myself the World: What Consciousness Is and How to Expand It.” For more about Koch’s research, check out this 2019 GeekWire article, which focuses on an earlier book titled “The Feeling of Life Itself.” You may also like my Fiction Science interview with Koch from 2021.

My co-host for the Fiction Science podcast is Dominica Phetteplace, an award-winning writer who is a graduate of the Clarion West Writers Workshop and lives in San Francisco. To learn more about Phetteplace, visit her website, DominicaPhetteplace.com.

Take a look at the original version of this item on Cosmic Log to get James Boyle’s recommendations for further reading, and stay tuned for future episodes of the Fiction Science podcast via Apple, Spotify, Player.fm, Pocket Casts and Podchaser. If you like Fiction Science, please rate the podcast and subscribe to get alerts for future episodes.

Three weeks after University of Washington biochemist David Baker won a Nobel Prize, the latest venture to spin out from his lab — Archon Biosciences — has emerged from stealth mode with $20 million in financing for a technology that uses computationally designed protein structures to treat cancer and other diseases.

The seed funding round was led by Madrona Ventures, with participation from DUMAC Inc., Sahsen Ventures, WRF Capital, Pack Ventures, Alexandria Venture Investments and Cornucopian Capital.

Archon’s proprietary protein structures, known as Antibody Cages or AbCs, have been years in the making. Archon’s CEO and co-founder, James Lazarovits, said the Nobel Prize that Baker won for his pioneering work in the field of protein design confirms his view that the newly unveiled startup is on the right track.

“It’s reaffirmed our conviction for why we’re in this place to begin with,” Lazarovits told GeekWire during a tour of Archon’s Seattle lab. “It’s doing things that were not possible before. … You could not do anything that we’re doing unless there was the convergence of all these different fields at this moment in time.”

The ABCs of AbCs

Archon’s AbC technology combines two molecular-based biomedical tools, naturally occurring antibodies and custom-designed proteins, to create a single new protein structure. These protein structures, or AbCs, are optimized with the aid of generative AI to travel in the body in controllable ways and engage with target cells in a more specific manner.

“There was almost no mystery in whether we could do it,” said George Ueda, Archon’s chief technology officer and co-founder. “It was just a matter of doing it.”

AbCs can be compared to Lego blocks, or to car frames, in the sense that different types of components can be mixed and matched for different purposes.

“At Volkswagen, if you want to make a mini-car or an SUV, they share a design framework known as MQB that shares parts and equipment for improved flexibility, cost savings and speed across multiple factories,” Lazarovits explained.

“By building a modular design platform, we readily iterate and optimize each design quickly, efficiently and inexpensively, based on data that comes in,” he said. “It’s a validated and efficient engineering framework that we employ at Archon to make better therapeutics faster. We can make two drugs for two different applications, and make it easier to do both.”

The custom-built nature of AbCs helps antibodies bind to the cells they’re supposed to target, while reducing side effects on cells that the AbCs are not intended to affect. “What we like to say is, it’s not whether you’re given a drug, it’s how you’re given it,” Lazarovits said. “How it behaves is actually incredibly important. … It’s not only how it’s distributed and retained, but also how it interacts with its target.”

Archon has already started to put AbCs through animal testing. “We’re printing off new AbCs every eight minutes,” Lazarovits said. “We have data to show we beat out a previous clinical molecule that failed to pass Phase II human trials, and we showed we were able to overcome the issues that limited the molecule’s success. And so, the AbC platform isn’t theoretical — we are making real products.”

This short Archon video shows how Antibody Cages are created:

Building a business

The work being done by Archon’s staff of 17 full-time employees has benefited from academic grants.

“As research faculty at the University of Washington, George and I were very on top of things with respect to grant funding,” Lazarovits said. “We earned / won over $7 million in grants, so we were in a position where we were essentially in the pre-seed stage. … We are asking these important questions that we know every investor would want to know, but we’re not doing it on anybody’s dime.”

A research paper laying out the concept behind Antibody Cages was published in the journal Science in 2021, and Ueda recalled discussing how to commercialize the technology with Baker, another one of Archon’s co-founders.

“David said, ‘Well, George, if we do this, no one’s going to develop it for you. … Have you ever thought about starting a company?'” Ueda recalled. “I said, ‘No, but I can’t not do this.’ There’s another shared belief that Jamie and I have — that if you have a unique opportunity, it’s on you to do it.”

Ueda and Lazarovits followed a path that’s much like the one taken by earlier spinouts from the University of Washington’s Institute for Protein Design. They licensed the AbC technology from UW and incorporated Archon Biosciences in 2023.

Even before Archon was founded, the researchers were on Madrona Ventures’ radar screen. “AbCs are a revolutionary approach to antibody therapeutics that creates immediately actionable engineering solutions to intractable drug development challenges,” Chris Picardo, who’s a partner at Madrona Ventures and one of Archon’s board members, said today in a news release.

“James and George have assembled a team of world-class scientists and pharmaceutical industry veterans that is able to create new preclinical molecules with extraordinary speed and accuracy,” Picardo said. “We’re excited to partner with them at the forefront of generative protein design and realize the myriad of possibilities to engage currently undruggable targets and pathways to improve human health.”

Taking aim at cancer

Archon is already focusing in on its first therapeutic applications, most likely including cancer treatment.

“We can either make a cancer cell die a lot, die a little bit, be neutral, or target an immune cell that fights the cancer and turn that up, activate that,” Ueda said.

Lazarovits said it’s too early to be more specific about the applications.

“There are very obvious applications in oncology,” he said. “We’re currently not going into explicit detail on the program, but we’re in a fortunate position. … Because we can insert ourselves so well into this established framework, we hope in the not-too-distant future that we’ll have these lead programs defined.”

The newly announced funding round is meant to take Archon to the next level.

“We will get to a point with this money that we will be able to clearly articulate our lead program, and how it offers differentiated value relative to therapeutics that have previously developed or deployed in a clinic,” Lazarovits said. “I would say that probably within the next 18 months, we’ll have strong conviction over lead programs to move into the clinic.”

A team led by researchers from the Pacific Northwest National Laboratory is finding new ways to accelerate the pace of computational chemistry, by making tools for quantum computing and AI-assisted data analysis available via the cloud.

Their effort to make supercomputer-scale resources more widely available through cloud computing could aid in the search for methods to break down toxic “forever chemicals” that are currently hard to get rid of. And that’s just one example.

The researchers describe their progress on the project — known as Transferring Exascale Computational Chemistry to Cloud Computing Environment and Emerging Hardware Technologies, or TEC⁴ — in a study published today in the Journal of Chemical Physics.

“This is an entirely new paradigm for scientific computing,” PNNL computational chemist Karol Kowalski, who led the cross-disciplinary effort, said in a news release. “We have shown that it’s possible to bundle software as a service with cloud computing resources. The initial proof of concept shows that cloud computing can provide a menu of options to complement and supplement high-performance computing for solving complex scientific problems.”

The TEC⁴ project makes use of resources including Berkeley Lab’s Perlmutter supercomputer, Microsoft’s cloud-based Azure Quantum Elements platform and NWChem, a computational chemistry software package developed at PNNL.

Study co-author Nathan Baker, product leader for Azure Quantum Elements, said Microsoft’s collaboration with PNNL “is a great example of how modern AI and HPC [high-performance computing] tools can advance computational chemistry.”

In their proof-of-concept exercise, the researchers ran a simulation of chemical interactions involving perfluorooctanoic acid — which is one of the industrial “forever chemicals” that was historically used for the production of non-stick coatings and firefighting foam. The chemical, also known as PFOA, was banned worldwide in 2019 but still persists in the environment.

The simulation traced a step-by-step process that broke down some of PFOA’s molecular bonds. “This sequence of reactions highlights a potential pathway for the degradation of PFOA, providing valuable insights into the chemical behavior and breakdown of this persistent environmental pollutant,” the researchers reported.

Now the researchers are recruiting additional collaborators to put their cloud-based platform to more rigorous tests. “We are building a family of codes,” Kowalski said. “The goal is to build a community around this effort.”

A graduate-level course that builds on the TEC⁴ project has already been developed at the University of Texas at El Paso, in collaboration with PNNL and Central Michigan University. Kowalski said TEC⁴ is pointing the way not only to new chemicals, but also to new ways of doing chemistry.

“We envision an ecosystem of use cases from low-tier to high-tier jobs that take advantage of GPU-based computing now being used extensively for artificial intelligence and machine learning applications,” Kowalski said. “We want to allow users to take advantage of different layers of compute, paying only for what’s needed and bundling software with compute access. This is the first step toward that future state.”

Kowalski and Baker are among 32 authors of “Electronic Structure Simulations in the Cloud Computing Environment,” the study published in the Journal of Chemical Physics. Other authors represent PNNL and Microsoft Quantum as well as Berkeley Lab, Argonne National Lab, Central Michigan University, the University of Texas at El Paso and the University of Washington.

University of Washington biochemist David Baker has won a share of this year’s Nobel Prize in chemistry for more than two decades of discoveries about the molecular structure of proteins — discoveries that have led to new medical therapies, new materials and new startups.

“I’m very, very excited about the future,” Baker, who is the director of the UW Medicine Institute for Protein Design, said today during a Seattle news briefing. “I think protein design has huge potential to make the world a better place, and I really do think we’re just at the very, very beginning.”

Baker shares the prize with Demis Hassabis and John Jumper of Google DeepMind, who have also pioneered computational techniques for predicting protein structure. They will be awarded their medals at a ceremony in Stockholm, Sweden, on Dec. 10.

The Royal Swedish Academy of Sciences said Baker “has succeeded with the almost impossible feat of building entirely new kinds of proteins.”

“His research group has produced one imaginative protein creation after another, including proteins that can be used as pharmaceuticals, vaccines, nanomaterials and tiny sensors,” the academy said in a news release.

Proteins serve as the foundation for biological processes, literally unlocking the workings of the cell. But determining the 3D structure of those molecular “keys” is a devilishly complex task. “Life could not exist without proteins. That we can now predict protein structures and design our own proteins confers the greatest benefit to humankind,” the Swedish academy said.

Baker and his colleagues developed an early protein design tool and made the code available through a consortium called Rosetta Commons. That code was leveraged into a game-like crowdsourcing campaign called Foldit, which enlisted hundreds of thousands of computer users to help unravel the twists and turns of proteins.

The advent of artificial intelligence tools accelerated the pace of protein structure prediction — and today, the Institute for Protein Design and Google DeepMind are recognized as the world’s leaders. Microsoft’s chief scientific officer, Eric Horvitz, noted on LinkedIn that “we made the decision to go all-in with David” five years ago and work together on accelerating the development of vaccines and drugs with AI.

Half of this year’s Nobel Prize in chemistry, which comes with a $1 million cash prize, will go to Baker. Hassabis and Jumper will share the other half.

Baker’s research contributed to the development of the world’s first computationally designed protein medicine, a vaccine for COVID-19 that was pioneered by colleagues at UW Medicine. UW says Baker holds more than 100 patents and has co-founded 21 biotech companies — some of which have already been acquired.

Takeda Pharmaceuticals bought one of the Institute for Protein Design’s spinouts, PvP Biologics, for $330 million. AstraZeneca paid $1.1 billion to acquire a different spinout, Icosavax, which develops synthetic vaccines that target naturally occurring viruses. Other spinouts associated with the institute include Cyrus Biotechnology, Sana Biotechnology, A-Alpha Bio and Xaira Therapeutics.

In 2020, Baker won the Breakthrough Prize in Life Sciences. “Being able to design proteins from scratch, to do exactly what you want to do rather than modifying what we find around us, is kind of like the transition out of the stone age,” he told GeekWire at the time.

Today, Baker said the technology could be harnessed to synthesize new treatments for maladies including cancer, Alzheimer’s disease and autoimmune disease; to produce chemicals that break down plastics and other pollutants; and to create new materials for applications that might be considered close to the lunatic fringe.

“You can imagine all kinds of hybrid materials that could be superior in many ways to any of the materials that we have today, including semiconductor materials,” he said. “So, it could be a new way of patterning things for electronics. I’d say that still qualifies as lunatic fringe.”

During a panel discussion in June about the Seattle tech ecosystem, Ken Horenstein, general partner at Pack Ventures — which invests in startups with ties to the UW — said Baker’s lab was “transformational to the world.”

“It is a matter of when he will win a Nobel Prize, not if he will win a Nobel Prize,” Horenstein said at the time.

In the wake of today’s announcement, David Younger, CEO of A-Alpha-Bio, called Baker a “brilliant and visionary biochemist.” Nikesh Parekh, who was CEO of Bio Architecture Lab, another startup spun out from Baker’s lab, said he’s “a true pillar of the Seattle tech community.”

“David is an inspirational and incredibly supportive academic co-founder to work with, in that he understands that both the opportunities and the challenges of commercialization,” Parekh told GeekWire in an email.

Baker is the eighth UW faculty member to receive a Nobel.

“Let’s be honest: This is about as good as it gets,,” UW President Ana Mari Cauce said at today’s news briefing. “We are so proud to be the home of some of the most remarkable, creative, innovative and talented faculty, and we don’t have a better example of that than David.”

Baker, 62, was born in Seattle and grew up not far from the UW campus. Both his parents were UW faculty members, and he graduated from Seattle’s Garfield High School in 1980. “He really is a local kid, and he ‘s having a global impact,” Cauce said.

Baker completed his undergraduate studies at Harvard in 1984, and earned his doctorate in biochemistry in 1989 at the University of California at Berkeley. He’s been a faculty member in the Department of Biochemistry at the UW School of Medicine since 1993. “I’ve been here for quite a long time, and I never thought for a nanosecond about leaving,” Baker said.

His wife, UW Medicine biochemist Hannele Ruohola-Baker, is a noted stem-cell biology researcher. Baker told The Associated Press that Ruohola-Baker was beside him when he got the news about the Nobel early this morning, and that she immediately started screaming.

“It was a little deafening,” Baker told AP.

GeekWire’s Taylor Soper contributed to this report.

The U.S. Coast Guard took a deep dive into the regulations governing submersibles today at a public hearing looking into the causes of last year’s loss of OceanGate’s Titan sub and its crew. And the issues raised sometimes got as murky as the depths of Puget Sound, where Titan underwent its first tests.

Among the witnesses who testified at the hearing in South Carolina was John Winters, the master marine inspector for Coast Guard Sector Puget Sound. For more than a decade, Winters worked with OceanGate CEO Stockton Rush on the regulatory requirements for two of the Everett, Wash.-based company’s subs, known as the Antipodes and Cyclops 1. But today he told the Coast Guard’s Marine Board of Investigation that he had nothing to do with Titan.

Winters recalled a time, about two years ago, when he was at OceanGate’s headquarters on the Everett Marina to check on one of the two submersibles in the Coast Guard’s records. He said he saw the three subs on a barge, and someone told him, “We finally got our submarine to go to the Titanic.”

“But that was the only thing in passing,” Winters said. “Nothing about what it was constructed to, who witnessed it. None of that stuff. Just, ‘Here it is, look at the outside.’ … That’s as far as it went.”

In the wake of the Titan tragedy, the Coast Guard is likely to go further. One of the objectives of this month’s hearings is to lay the groundwork for regulatory changes that would help head off future fatal incidents involving submersibles.

Five people died when Titan underwent a catastrophic implosion during its descent to the Titanic in June 2023. In addition to Rush, who was piloting the sub, the victims included veteran Titanic explorer P.H. Nargeolet, British aviation executive Hamish Harding, Pakistani-born billionaire Shahzada Dawood and his son, Suleman.

The investigative board has already heard witnesses recount their concerns about OceanGate’s use of carbon-fiber composite for its hull. Two witnesses — from NASA and Boeing — added context to those concerns today. But most of the hearing was devoted to the regulatory gaps that the Titan tragedy brought to light.

Winters said Rush made no effort to get Titan inspected by the Coast Guard. And although OceanGate worked with the Coast Guard to get its other two subs received designations as oceanographic research vessels, or ORVs, Winters said Rush often complained that “regulations were stifling his innovation process.”

Small submersibles that are used exclusively as research vessels are exempt from some Coast Guard regulations — including one that requires inspections for subs that take on paying customers. OceanGate’s business model relied on having “mission specialists” pay a fee to participate in what were portrayed as research missions.

“Is it an ORV? Then, are they scientific personnel? If both of those answers are yes, the payment matter is moot,” said Lt. Cmdr. Jonathan Duffett of the Coast Guard Office of Commercial Vessel Compliance.

It’s typically up to the officer in charge of marine inspection in each Coast Guard sector to determine whether a sub and its crew are involved in a research mission. Winters said communication between sectors can sometimes get spotty. He said he would receive notices about Antipodes’ missions in other jurisdictions, but didn’t get notified about OceanGate’s Cyclops 1 dive into New York’s Hudson Canyon in 2020.

The Cyclops 1 dive to the wreck of the Andrea Doria in 2016 raised even more questions: Investigators said Cyclops didn’t have its ORV designation for that voyage, but took on a mission specialist who paid $35,000 for the trip.

Duffett ran through all the requirements for small submersibles, and then said that “if all those boxes are checked, then, yes, it would fall under the definition of a small passenger vessel and have to have a certificate of inspection.”

And then there’s Titan. The fact that Titan went out to sea from a Canadian port and operated in international waters for its Titanic dives, the Coast Guard didn’t exercise any oversight over those dives. But could the Coast Guard have played more of a role when Rush was testing Titan in Puget Sound? That was the question on the mind of Jason Neubauer, chair of the Coast Guard Marine Board of Investigation.

“If OceanGate were testing a pressure submersible with just the owner on board, or a crew member, before the vessel was certified or registered, could they legally do that?” he asked Duffett.

Duffett hemmed and hawed a bit. “As far as Coast Guard requirements go, as long as they’re in compliance with Subchapter C, and there are other generally applicable requirements, environmental requirements and 33 CFR rules of the road, that sort of thing … but it’s feasible to do that legally. Yes.”

The Coast Guard is due to wrap up its hearings on Friday with testimony from former OceanGate employee Matthew McCoy; Capt. Jamie Frederick of Coast Guard Sector Boston, who helped lead the search effort for the Titan sub; and Scott Talbot, a search-and-rescue specialist at the Coast Guard.

Other highlights from the hearing

Boeing provided engineering advice to OceanGate at the very start of the design process for the Titan sub, Boeing engineer Mark Negley told the investigative board. He said OceanGate got in touch with Boeing in the 2013 time frame because of the supporting role that Boeing played in the development of an autonomous underwater vehicle known as Deepglider for the University of Washington.

Negley said Boeing engineers looked at the feasibility of using carbon-fiber composite for the submersible’s hull, and noted the challenges that would be involved in building up a cylindrical shell thick enough to meet OceanGate’s specifications. Boeing also helped OceanGate record acoustic readings during tests of subscale hull models. But the relationship fizzled out in the 2016-2020 time frame — and Negley was asked why.

“We had a number of different requests from OceanGate to respond to different requests for proposals,” Negley said. “I don’t know exactly. I think maybe we were too expensive.”

In June, Wired magazine reported that Negley sent an email to OceanGate CEO Stockton Rush in 2018, warning him that Titan’s carbon-fiber hull had “a high risk of significant failure at or before you reach 4,000 meters.” Negley reportedly included a graph showing how the strain would increase at lower depths, with a skull and crossbones marking the region below 4,000 meters. That email didn’t come up during today’s hearing.

NASA had an agreement with OceanGate in 2020 to help the company manufacture the carbon-fiber hull that the Titan sub used for Titanic trips. NASA engineer Justin Jackson confirmed that the contract called on OceanGate to pay $148,874 for engineering advice as well as production, testing and analysis of scale models of the hull. Jackson said NASA was interested in the project because the technology could be used to build space habitats and radiation shields.

The partnership didn’t last long. “The COVID pandemic occurred, and we weren’t able to accomplish any of the fabrication. We provided remote consultation throughout the build of their one-third-scale article, but we did not do any manufacturing or testing of their cylinders,” Jackson said. “We received roughly $40,000 for the remote consulting effort during the one-third-scale build. We returned close to $124,000 of that $148,000.”

Jackson said OceanGate tried to maintain its connections with NASA. “They wanted to do another press release that we ultimately couldn’t come to a resolution on the details of, and then conversations ended shortly thereafter,” he said.

He was asked why NASA nixed the news release. “It was the language they were using,” Jackson said. “It was getting too close to us endorsing. Our folks had some heartburn with the endorsement level of it.”

Previously:

• NTSB spots flaws in Titan sub’s hull and focuses on a bang
• Submarine expert recounts how a Titan trip unnerved him
• Former top engineer says cost concerns shortchanged safety
• OceanGate client recalls a tricky Titanic tangle
• Titanic traveler makes a tearful plea for citizen science
• Videos show the shattered remains of Titan sub
• OceanGate whistleblower traces the roots of his concerns
• Hearing reveals the last words sent by Titan’s crew
• A new chapter in the OceanGate probe, but not the end

A pair of reports by the National Transportation Safety Board found evidence of imperfections in the carbon-fiber hull that was made for OceanGate’s Titan submersible — plus indications that the hull behaved differently after a loud bang was heard at the end of a dive in mid-2022.

At the time, OceanGate team determined that the loud bang was not a serious problem, but less than a year afterward, the sub and its crew were lost in a catastrophic implosion during a trip to the wreck of the Titanic in the North Atlantic,

Donald Kramer, an senior materials engineer who presented NTSB’s findings today at a Coast Guard hearing in South Carolina, declined to go beyond the data and speculate on whether the imperfections or the bang figured in Titan’s doom. But one leading theory for the sub’s failure suggests that weaknesses in the hull gave way under the extreme pressure of the deep ocean.

The five crew members lost in last year’s tragedy included Stockton Rush, who was the CEO and co-founder of Everett, Wash.-based OceanGate as well as the sub’s pilot; veteran Titanic explorer P.H. Nargeolet; British aviation executive and adventurer Hamish Harding; and Pakistani-born business executive Shahzada Dawood and his son, Suleman.

The NTSB based its newly published reports on a study of debris that was brought up from the seafloor, plus an examination of carbon-fiber samples that were cut off the ends of OceanGate’s hulls.

Two hulls were fabricated: The first one was used for test dives in 2018 and 2019 in the Bahamas, but was scrapped when a crack was discovered in the 5-inch-thick carbon-fiber composite. The second one was used for Titanic dives in 2021, 2022 and 2023. In addition, the NTSB studied samples from subscale test hulls.

Kramer said the surplus pieces from Titan’s second hull had “several anomalies within the composite and the adhesive joints, including waviness, wrinkles, porosity and voids.” Bulges in the hull were ground smooth at 1-inch intervals during the process of laying down the carbon-fiber strands, according to the reports.

Debris from the sub’s hull showed delamination of the composite, primarily at places between the 1-inch layers.

In addition to looking at the carbon-fiber hull samples, the NTSB analyzed data captured by acoustic sensors and strain gauge sensors mounted in the hull. OceanGate used the data from those sensors in a real-time monitoring system that it had hoped would provide advance warning of problems with the hull.

The NTSB focused on data from a notable dive to the Titanic in July 2022. At the end of that dive, the crew heard a loud bang resonating through the hull. One of the mission specialists who paid to participate in the dive, Fred Hagen, said during earlier testimony that “you would have to be brain-dead not to be somewhat concerned” about the noise.

According to other testimony, reports of the bang also stirred concern among other OceanGate team members at the time — including Phil Brooks, who was the director of engineering, and tech contractor Antonella Wilby.

Hagen said OceanGate determined that “the body of the fuselage of the Titan had just jumped in its carriage, so there was no damage.” Brooks said that “seemed like a reasonable explanation” at the time. But based on its study of the sensor data, the NTSB said strain gauges placed on a specific section of the hull showed significant differences in readings taken before and after the bang occurred.

“It means that there is a sudden change in both the longitudinal and the hoop strain at this particular location of the hull. … There’s a change in the strain in the hull that is occurring at this location that is not occurring at other locations,” Kramer said.

There were also unusual spikes in acoustic activity during the first dive that took place after the bang was heard, but the second and third dives afterward “show a trend of the hull becoming quieter,” Kramer said.

Kramer emphasized that the NTSB was continuing its tests and its analysis of the results. He dodged a question relating to causal connections between Titan’s implosion and the issues that were detailed in the agency’s reports. “That is still subject to our own internal analysis at this point,” Kramer said.

Eventually, the NTSB and the Coast Guard will produce separate reports delving into the causes of the Titan tragedy, and providing recommendations for regulatory changes aimed at heading off such tragedies in the future.

Other highlights from the hearing

William Kohnen, CEO and founder of Hydrospace Group, said the viewport that his company provided for OceanGate’s first hull was rated for use at a maximum depth of 2,150 feet (650 meters), but not for the Titanic depth of 12,600 feet (3,840 meters). He said he pleaded with the OceanGate team to get a better window. “We figured they could buy it in any number of places, and in the end, they did. They just didn’t buy it from us,” he said.

The investigative board’s chair, Jason Neubauer, said the viewport for the second hull was purchased in 2020 from a German company called Heinz Fritz.

Kohnen was the author of a controversial draft letter that questioned OceanGate’s practices in 2018. The letter was supposed to have been sent to OceanGate CEO Stockton Rush by the Marine Technology Society, but the MTS board told Kohnen it couldn’t be issued on the society’s behalf.

“It was considered outside the bylaws of the nonprofit organization,” Kohnen recalled. “I had to respect that, and I said OK, and I filed the letter. Was a letter received? That I can confirm, it was, because somewhere down in all the electronic letters that were circulated, one of them got passed on to OceanGate, and Stockton got the letter. And Stockton’s first reaction was to call me.”

Kohnen said Rush argued against his suggestion that OceanGate should get the Titan sub certified by outside experts. “It was the usual response that it takes too long, it’s too expensive, and they don’t know about this technology,” he recalled.

“I don’t think many people ever told Stockton ‘no,'” Kohnen said. “I don’t think he understood that concept very much.”

Bart Kemper, principal engineer at Kemper Engineering Services, said the root cause of the Titan sub implosion was still “indeterminate” — but he outlined potential failure modes involving the carbon-fiber hull, the bond between the hull and the sub’s titanium end caps, and the sub’s acrylic viewport.

In answer to a question that referred back to Kramer’s testimony, Kemper said that delamination in a section of the carbon-fiber composite hull could cause a shift in strain readings. He said he would not have gone ahead with further dives if a significant change in strain was detected in the hull. “I would have stopped operations and assessed,” Kemper said.

Kemper also criticized OceanGate’s engineering approach for not doing a life-cycle analysis of the hull. “This is just wrong,” he said. “It’s flat wrong.”

He said authorities should limit the crew of an experimental craft to an operator and assistant, with no passengers or cargo. He recommended that submersibles should be required to have a minimum level of navigation aids, emergency gear and a standardized lifting point for extraction by a remotely operated vehicle. Kemper said the Coast Guard and the submersible community should figure out “a reasonable minimum standard to require, and why and how to do it.”

Previously:

• Submarine expert recounts how a Titan trip unnerved him
• Former top engineer says cost concerns shortchanged safety
• OceanGate client recalls a tricky Titanic tangle
• Titanic traveler makes a tearful plea for citizen science
• Videos show the shattered remains of Titan sub
• OceanGate whistleblower traces the roots of his concerns
• Hearing reveals the last words sent by Titan’s crew
• A new chapter in the OceanGate probe, but not the end

Karl Stanley has done more than 2,000 dives in submersibles that he’s built himself, so he was intrigued when OceanGate CEO Stockton Rush invited him down to the Bahamas for a test dive in the company’s Titan sub in 2019. But by the end of that dive, Stanley’s curiosity turned to concern.

“In retrospect, there were a lot of red flags,” Stanley said today during a Coast Guard hearing into last year’s loss of the Titan sub and its crew as it was descending to the wreck of the Titanic in the North Atlantic. Rush, who was the sub’s pilot, and four other people lost their lives in last year’s tragedy.

The Coast Guard Marine Board of Investigation, meeting in South Carolina, is looking into the causes of last year’s tragedy and is expected to recommend measures to head off such tragedies in the future. Part of its job is to review OceanGate’s missteps during Titan’s development.

Back in 2019, Everett, Wash.-based OceanGate was testing the sub in the Bahamas, in anticipation of taking its first trips to the Titanic later that year. Rush had made a solo trip to more than 3,900 meters (12,800 feet) in depth the previous December, and Stanley was one of three people set to accompany him on a follow-up dive.

Stanley, who operates a submersible tour company in Honduras, had come to know Rush through the tight-knit community of sub operators. He recalled traveling to Everett to help out with the construction of Titan’s landing and recovery platform — and said he was “excited” about OceanGate’s plan to use a lightweight carbon-fiber hull for Titan.

For the test dive, Titan was brought about 10 miles offshore, and made a gradual descent into the depths. Stanley said he had been warned to “be prepared for noises” — but even with that warning, he wasn’t prepared for what he heard on the way down.

“It’s making noises,” he recalled. “The sound of a carbon-fiber band breaking in a 5-inch matrix is a lot of energy being released. We were all clearly a little disnerved, but he had warned us, and I think nobody wanted to be the one to say they wanted to go up first.”

Stanley said the cracking sounds got louder as the sub neared the Titanic depth of roughly 12,500 feet. “At some point, I don’t remember who — we were all like, well, that’s probably close enough. We’ve been down here long enough,” he said.

“There was kind of like a grand finale of cracking sounds as we were getting close to the surface,” Stanley recalled. He speculated that the sounds were caused by a release of energy that was stored in the carbon fibers during exposure to the extreme pressure in the deep ocean.

Stanley was so shaken by the experience that he voiced his concerns about the hull in a weeks-long email exchange with Rush. “The only question in my mind is, will it fail catastrophically or not,” he wrote in one email.

Stanley argued that Titan should undergo at least 50 tests before paying clients were allowed on board. Rush insisted that OceanGate’s testing routine was sufficient to ensure safety — but Stanley also got the message that he shouldn’t talk to others about the cracking sounds he heard.

Although Stanley agreed to keep mum, he told Rush in a follow-up email that he wasn’t comfortable about it. “The fact that you indirectly told me to not speak about the noises I heard on the dive, to me says a lot,” he wrote.

“I kept trying to tell him, ‘You do not have a marketable product,'” Stanley said at today’s hearing.

Stanley said he didn’t know at the time that the Titan sub had been struck by lightning during an earlier series of tests, or that OceanGate’s engineers had identified a crack in the hull a month and a half after his trip. OceanGate canceled that year’s planned Titanic expeditions. However, the company blamed the cancellation on regulatory problems having to do with a support ship — and never mentioned the crack in the hull.

In 2020, OceanGate ordered a new hull to be manufactured by two Seattle-area companies, ElectroImpact and Janicki Industries. In 2021, the company began conducting Titanic expeditions with the replacement hull. Titan’s second hull gave way in June 2023, killing Rush as well as veteran Titanic explorer P.H. Nargeolet, aviation executive and adventurer Hamish Harding, Pakistani-born business executive Shahzada Dawood and his son, Suleman.

One leading theory for the carbon-fiber hull’s implosion is that repeated trips to the Atlantic’s depths, plus exposure to the elements between trips, weakened the hull or the seals between the hull and the sub’s titanium end caps.

If Stanley knew everything he knows now, would he have gone on that test dive five years ago?

“There’s a lot of things that, if I had known, I wouldn’t have gone,” he said. “People have told me that I was, you know, stupid, naive. But really what it came down to was, at that point, I had no reason to believe that Stockton was a liar, and I had no evidence of any lies on his part. When he said that he did a proper amount of testing — he’s a trained engineer, he’s not an idiot.”

Stanley’s perspective has changed since then. “A lot of the things I’ve learned in the last 18 months have been very shocking,” he acknowledged.

Other highlights from the hearing

Did Stockton Rush have a death wish? Stanley said his evolving view on “Stockton’s psychology” led him to believe that Rush went ahead with the dives to the Titanic — even though he knew about the problems with the sub — because he wanted to “leave his mark in history,” and because he was feeling pressure from OceanGate’s investors and clients. Stanley noted that Rush came from a famous family, with two signers of the Declaration of Independence among his ancestors.

“He knew that eventually it was going to end like this, and he wasn’t going to be held accountable, but he was going to be the most famous of all his famous relatives,” Stanley said.

OceanGate’s staff shrunk from about 30 employees in 2019 to 14 or 15 employees in 2023, said Amber Bay, who served as the company’s director of administration during that time frame. Bay said the COVID pandemic was a factor behind the cutbacks. “Any kind of operations at sea were then halted, obviously getting into a submersible and onto ships,” she said. “So we had to get rid of many of our operational staff … because they were no longer needed.” After the pandemic, “we were still in recovery mode.”

Bay confirmed that Stockton Rush occasionally had to put more of his own money into OceanGate’s account to meet payroll. “He increased his investment by making a deposit,” Bay recalled. When asked to identify the main revenue sources supporting OceanGate, Bay replied: “Investors.”

She also confirmed that OceanGate asked employees to delay getting paychecks at the beginning of 2023, because “the finances were just getting very tight at that time.” She said she deferred her own pay, as did Rush.

Bay was asked about the dismissal of Antonella Wilby, an employee who raised concerns about a loud bang that was heard at the end of a Titan dive in 2022. Bay said that Rush fired Wilby because she had “behaved erratically and disturbed the crew.” She said she didn’t recall telling Wilby that she lacked an “explorer mindset,” which was something Wilby said in her testimony last Friday.

In the wake of the accident, OceanGate suspended all exploration and commercial operations.

Bay concluded today’s testimony with an tearful statement about the tragedy. “I had the privilege of knowing the explorers whose lives who were lost — Stockton, P.H., Shahzada, Suleman and Hamish — and there’s not a day that passes that I don’t think of them, their families and the loss,” she said. “It’s been a difficult year for them, and for all of us.”

Previously:

• Former top engineer says cost concerns shortchanged safety
• OceanGate client recalls a tricky Titanic tangle
• Titanic traveler makes a tearful plea for citizen science
• Videos show the shattered remains of Titan sub
• OceanGate whistleblower traces the roots of his concerns
• Hearing reveals the last words sent by Titan’s crew
• A new chapter in the OceanGate probe, but not the end

OceanGate’s Titan submersible briefly became tangled up in the wreck of the Titanic during a 2022 dive, a mission specialist who was on the sub told investigators today.

“We had a skid stuck for a minute,” Fred Hagen said during a hearing in South Carolina that focused on the causes of last year’s loss of the sub and its crew. “It wasn’t a big deal.”

The Coast Guard’s Marine Investigation Board is reviewing the history of Everett, Wash.-based OceanGate’s Titan sub development effort, with the aim of making recommendations to avoid future undersea tragedies.

Last year’s catastrophic implosion killed five people: OceanGate CEO Stockton Rush, who served as the sub’s pilot; veteran Titanic explorer P.H. Nargeolet; British aviation executive and adventurer Hamish Harding; and Pakistani-born business executive Shahzada Dawood and his son, Suleman.

Hagen went on two Titan dives — one in July 2021, which was aborted when one of the sub’s thrusters malfunctioned, and the other in July 2022, which successfully reached the Titanic at a depth of 12,600 feet (3,840 meters).

Nargeolet steered the sub as the crew members took in the shipwreck’s iconic sights, including the bow of the 112-year-old wreck and the ruins of the Grand Staircase. But Hagen said he wanted to see more, and he persuaded Nargeolet to head back toward the ship’s stern section. “I’d asked him to go around where the break was, and for a few moments we had gotten stuck,” he said. “He was very quiet, and he was working the controls. … I leaned over, and I said, ‘P.H., it seems that we’re stuck.’ And he says, ‘Yes, Fred, we are.'”

Hagen said that the skid was momentarily snagged in “pipes and things” on the Titanic wreck, but that Nargeolet managed to free up the sub after no more than a minute or two. The surface support team became concerned about what was happening and “told us to come up immediately,” Hagen said.

“Obviously, when you’re down there, it feels like a big deal. I think P.H. certainly wasn’t overly concerned,” he said.

Hagen’s tale got the panel’s attention. Keith Fawcett, technical adviser to the Coast Guard Marine Board of Investigation, asked him whether OceanGate had alerted the Canadian or U.S. Coast Guard about the Titanic dive in advance, so that they could have been ready to provide assistance if the sub became entangled in wreckage.

“No,” Hagen replied. “But I did have clear conversations with at least P.H., and probably others. The conversation, as related to me, was that there were few assets on Earth capable of getting to depth, and that if something went wrong, we were all going to die.”

An attorney for OceanGate, Jane Shvets, asked whether the entanglement with the wreck was intentional on OceanGate’s part. “No, it certainly wasn’t intentional,” Hagen said. “It was completely my fault, nobody’s fault but mine. I was egging P.H. on, but it was not intentional, and it really was not a big deal.”

For what it’s worth, U.S. law prohibits people subject to U.S. jurisdiction from disturbing the Titanic wreck site without government authorization, and OceanGate’s stated policy was that it would not disturb the site. During a Reddit “Ask Me Anything” session in 2020, Stockton Rush said “we are not disturbing the wreck and just documenting it.” He acknowledged, however, that the currents in the deep ocean were “clearly challenging.”

Hagen, who owns a construction company in Pennsylvania, said he and the other Titan crew members accepted the risks that were involved in diving to the Titanic.

“You don’t do it because it’s safe,” he said. “You do it because of the adrenaline rush. … It was not a safe environment in the Titan, and it was never supposed to be safe. If it was safe, then you might as well just get in a trolley car and ride around town.”

Other highlights from the hearing

Hagen confirmed reports that a loud bang was heard when the sub was surfacing at the end of the 2022 dive — throwing a scare into the crew. “You would have to be brain-dead not to be somewhat concerned,” he said. But Hagen said those concerns were assuaged once the crew was brought out and the sub was inspected. “It turned out, after we got on the ship, that the body of the fuselage of the Titan had just jumped in its carriage, so there was no damage,” he said.

Hagen’s 2021 dive was also marked by anomalies. He said the dive had to be aborted because the sub was off-balance and started spiraling off course. When the crew tried to turn on the sub’s thrusters for a course correction, the starboard thruster failed to activate. It also took longer than expected to jettison the weights for the sub’s ascent.

When the sub resurfaced and was hoisted back onto its mothership, “it slammed down on the deck with quite a bit of force,” Hagen said. “Now, the complicating factor was that the decision had been made to only install four of the 18 bolts in the 3,500-pound titanium dome.” That decision was aimed at reducing the amount of time required to open up the dome and bring out the crew at the end of the mission, Hagen said.

The force of the sub hitting the deck caused those four bolts to shear off. “They shot off like bullets, and the titanium dome fell off,” Hagen said. After that, all 18 bolts were installed for every mission.

OceanGate’s engineering team was “a tough group to work with,” said Dave Dyer, an engineer at the University of Washington’s Applied Physics Laboratory. APL-UW provided engineering services during development of OceanGate’s Cyclops 1 submersible and in the early stages of the Titan project, starting in 2013. But the relationship cooled in the 2017-2020 time period, due to differences over issues including the potential use of glass housings on the exterior of the hull, and the integration of the carbon-fiber hull and the titanium end caps. “We were butting heads too much over that,” Dyer said.

APL-UW partnered with OceanGate in 2016 to test a one-third-scale model of the hull for Titan, which was originally known as Cyclops 2. Dyer said OceanGate continued to use a pressure chamber at UW for testing, but APL wasn’t involved in those follow-up tests.

At the time that APL-UW backed away from the Titan project, it looked as if the Titan sub’s design was “heading down the right path,” Dyer said. But he acknowledged that he “was really questioning whether they could do it.”

Dyer said APL-UW also helped OceanGate develop an acoustic monitoring system for the carbon-fiber hull. He said the plan, as he understood it, was to use such a system to determine whether the hull should undergo further testing or replacement before a subsequent dive. “We were under the impression that they were not planning to use it for real time,” Dyer said.

Certification should become a requirement for submersibles, said Triton Submarines CEO Patrick Lahey. OceanGate did not make the effort to certify the Titan sub, and argued that the process of certification, or “classing,” was too cumbersome to match the pace of innovation. But Lahey, whose company has built about 30 submersibles for its customers, said the need to require certification was “the most important takeaway from all this.”

“Yes, there’s additional cost, there’s additional time and hassle associated with it, but the end result is you’ve got a piece of machinery that meets an internationally recognized set of standards and that people can use with confidence, knowing that an independent peer review of a third-party classification society has been conducted on it,” he said.

OceanGate tech contractor Antonella Wilby told the board that the company used an “idiotic” navigation system that required coordinates to be transferred manually from one system to another. “The primary navigational map was a hand-drawn map that showed the bow of the Titanic and the debris field and the stern,” she said. “You could get a lot of error just by accidentally moving this map a few pixels off the reference point.”

Wilby said her teammates on the support ship weren’t all that supportive when she suggested streamlining the navigation system with different software, or when she expressed concerns about the loud bang that was heard toward the end of Hagen’s mission. She said she thought about escalating her concerns to OceanGate’s board of directors, but was told by a colleague that she risked running afoul of her nondisclosure agreement. “She cautioned me that the company is extremely litigious, and that matched with certain things I’d heard,” Wilby said.

When Wilby was asked whether she thought OceanGate was operating safely, she said no. “No aspect of the operations seemed safe to me,” she said.

Previously:

• Titanic traveler makes a tearful plea for citizen science
• Videos show the shattered remains of Titan sub
• OceanGate whistleblower traces the roots of his concerns
• Hearing reveals the last words sent by Titan’s crew
• A new chapter in the OceanGate probe, but not the end

Amateur adventurer Renata Rojas took a trip to the Titanic in OceanGate’s Titan submersible in 2022, and she was aboard Titan’s support ship last year when the sub and its crew were lost. Now she’s worried that the regulatory response to the tragedy will close off opportunities she and other citizen explorers have enjoyed.

Rojas fought back tears as she shared her concern today at a Coast Guard hearing that was aimed at determining the cause of Titan’s loss and formulating recommendations to avoid future tragedies.

“What we’ve all gone through is still very raw. Nothing is going to bring our friends back,” she told the investigators on the panel.

“I hope that this investigation creates an understanding that with exploration, there’s risk. And without taking that risk and the exploration, the world would still be flat,” she said. “I hope that innovation continues so we can make the oceans accessible to people like me who got to fulfill a dream, and that you still allow citizen scientists to participate in expeditions.”

Rojas is a Mexican-born, New York-based banker who began diving when she was 11 years old and long dreamed of seeing the Titanic up close. She participated in OceanGate’s dive to the wreck of the Andrea Doria in 2016, and also took submersible trips through New York’s Hudson Canyon and in the waters close to OceanGate’s home base in Everett, Wash. All that led up to her Titanic dive.

The price tag was more than $100,000, but Rojas said she accepted the cost — and the risk. “Problems arise,” she said. “This was never really sold as a Disney ride. This was an expedition where things happen, and you have to adapt to change.”

She recalled one trip — not her own — where Titan’s crew “spent the night in the sub” because rough weather temporarily ruled out a timely return to the surface. “They slept,” she said.

In June 2023, Rojas served as a volunteer on the Polar Prince, the support ship for Titan’s final dive. She said she assisted the crew members as they entered the submersible in preparation for the expedition.

“”They have to take their hat out, their life jacket out. We all place that into each individual bag so they know it’s their belongings,” Rojas explained. “I was the one holding that bag for them as they got in the sub.”

She recalled that Titan’s last crew was “happy to go.”

“I saw five people smiling, looking forward to their journey,” Rojas said. “We had wonderful weather. We had been waiting for a while for the wonderful weather. Just everything was done on time. So, we’re all excited.”

The five people making the dive were Stockton Rush, who was OceanGate’s co-founder and CEO as well as the sub’s chief pilot; veteran Titanic explorer P.H. Nargeolet; British aviation executive and adventurer Hamish Harding; and Pakistani-born business executive Shahzada Dawood and his son, Suleman.

Rojas recalled that all seemed to go well until contact with the sub broke off about an hour and a half into the dive. The support team tried to restore the communication link — but after hours of silence, the Coast Guard and other authorities were notified. “We went into ‘go’ mode at that point,” Rojas said.

Four days later, a remotely operated vehicle captured views of the wreckage from Titan’s catastrophic implosion — confirming the loss of the crew.

This month’s hearings are expected to result in tighter regulations for submersible trips, on the federal level and perhaps on the international level as well. In her closing statement, Rojas said she hoped the Coast Guard and other maritime authorities would recognize the need for the kind of privately funded deep-sea exploration that served as the basis for OceanGate’s business model.

“Expeditions are needed, and the oceans are significantly underfunded,” she said. “Private citizens are the ones funding the expeditions, and I hope that doesn’t stop.”

Among other highlights from today’s hearing in South Carolina:

• Rojas took issue with OceanGate whistleblower David Lochridge’s description of the Andrea Doria expedition in 2016. Lochridge suggested that Rush was reckless in his piloting of the Cyclops 1 submersible, and that a heated argument broke out over who should take the controls. “He must have gone on a different dive,” Rojas said. “Nobody was panicking, nobody was crying, and there was definitely no swearing and yelling.”
• Steven Ross, a marine biologist who served as OceanGate’s chief scientist for the Titanic expeditions, acknowledged that he was in the sub last year for a rough dive that preceded the final, fatal trip. That outing, which didn’t go to the Titanic, ended in a tumble when a malfunction on the sub’s platform left Titan stuck at a 45-degree slant. “The pilot crashed into the rear bulkhead,” Ross said. “The rest of the passengers tumbled about. I ended up standing on the rear bulkhead. One passenger was hanging upside down, and the other two managed to wedge themselves into the bow end cap.” He said the ordeal lasted about an hour, but no one was injured.
• Ross said that he managed the scientific program for OceanGate, and that a research paper based on data gathered during Titan’s dives was “in development still.” But he acknowledged that OceanGate had “multiple agendas” for its expeditions. What did he consider the primary mission for Titan’s trips in 2021, 2022 and 2023? “To visit and view the Titanic shipwreck,” Ross replied.
• Ross said he once asked Rush how many dives he thought the Titan sub would be able to take on before it needed refurbishment. One of the panel members asked what Rush’s answer was. “Indefinite,” Ross recalled.

The Coast Guard hearings continue on Friday with testimony from OceanGate mission specialist Fred Hagen; Dave Dyer of the University of Washington’s Applied Physics Laboratory; and Patrick Lahey of Triton Submarines. The proceedings will be livestreamed beginning at 8:30 a.m. ET (5:30 a.m. PT):

Previously:

• Videos show the shattered remains of Titan sub
• OceanGate whistleblower traces the roots of his concerns
• Hearing reveals the last words sent by Titan’s crew
• A new chapter in the OceanGate probe, but not the end