octave van den Bosch retweeted
SpaceX has arrived.
We announced an agreement to acquire a nationwide low-band spectrum license portfolio that will pave the way for @Starlink to become a major mobile carrier in the US.
With this new spectrum and our Gen2 constellation, Starlink Mobile can ensure Americans have access to high-speed mobile broadband no matter where they are → spacex.com/updates#starlink-…
octave van den Bosch retweeted
Today we announced the Claude-led discovery of a molecular machine that we suspect could represent a new gene editing mechanism. Its precise function, biotechnological utility (if any), or level of significance is not yet clear, but at minimum it is work I would have been proud to do as a PhD student. The work was done mostly, though not entirely, by Claude: our life sciences team suggested a broad area of research, Claude read through the literature and a bunch of genome data and discovered something interesting, then Claude proposed experiments to verify the discovery and our team carried them out.
It’s easy to dismiss this as a one-off or curiosity, but we’ve repeatedly seen a pattern where AI performance in new intellectual domains goes from weak to superhuman in a matter of a few years. In 2023 models struggled to do math at the level of an average high-school student. In 2024 they started to do well on math competitions for the best high-schoolers in the country, in 2025 they started to solve minor open problems, in early 2026 more significant open problems, and in late 2026 they are beginning to solve the top few open problems in all of mathematics. We believe AI for biology is on a similar exponential trend.
The main difference between biology and mathematics, of course, is that math can be done purely theoretically, while biology requires experimentation. Some have used this to draw the conclusion that AI’s utility in biology will be limited. We think this is wrong. As we’ve demonstrated today, humans can collaborate with AI to perform the experiments, validate key results in a few weeks and, if necessary, work with the AI to iterate on what they find. Eventually it may even be possible for Claude itself to safely perform the experiments by autonomously controlling lab equipment, with appropriate safeguards in place, but we aren’t doing that today (our lab is also a BSL1/BSL2 facility that doesn't handle materials dangerous to humans).
More broadly, biomedical advancement has many stages — from fundamental biology discoveries, to translational research, to drug discovery, clinical trials, and finally the actual delivery of medicines and health care to patients. We are also interested in these later stages, but even simply accelerating the first stage of fundamental biological discoveries has the potential to speed up and broaden the entire pipeline. Improving our understanding of biology and sharpening biologists’ tools can drive forward all of the later stages, for example by identifying new drug targets, finding new therapeutic modalities, allowing for more precise measurement, and speeding up the experimental loop which itself further accelerates our understanding of biology. This will not in itself speed up clinical trial times, but if it succeeds it could greatly increase the number of promising candidates that go into the pipeline — an increase in throughput even though latency remains.
In Machines of Loving Grace, I wrote about AI’s potential to “cure most diseases in 5-10 years” — a goal that sounds impossible, but one I believe is just barely possible if AI is applied to every stage of the pipeline. The first step is showing that AI can first help with, and then drive, biological discoveries.
Claude’s discovery is the latest in a line of related prior work that goes back decades, beginning with systems like CRISPR, and continuing with discoveries like the bridge recombinase and VIPR in the past few years. Recently, there has been heightened interest in systems based on reverse transcriptase (RT) enzymes, the enzyme underlying the system Claude identified. And most recently, a Stanford team working independently described a novel RT system with an associated non-coding array that is in some ways similar to the one Claude found, though they are distinct systems that evolved independently from each other. I believe that we’re at the very beginning of finding such systems and developing them into powerful tools for biotechnology.
I’m proud of the resources Anthropic has invested in accelerating the public benefits of AI through the life sciences, and we’re aiming both to grow our life sciences team and to work with other scientists to extend this approach to a broad range of problems. If you have a proposal for a research collaboration or are interested in joining our life sciences team, please reach out.
Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR.
We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use.
Read more: anthropic.com/news/claude-di…
octave van den Bosch retweeted
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.
The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.
The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
octave van den Bosch retweeted
This is so true and so brilliant,
why do bad ideas originate in academia - because it’s the only place you are shielded from dealing with reality
octave van den Bosch retweeted
Moderna/Merck just ran a 1,137-patient Phase 3 trial where every single dose was unique to that patient's tumor. It worked.
The pipeline: surgical resection → whole exome + RNA sequencing → ML neoantigen ranking → mRNA encoding up to 34 patient-specific targets → manufactured and shipped in 8 weeks. One drug, different sequence for every patient.
The ML step is worth to note: the algorithm ingests WES + RNA-seq to identify somatic mutations, then predicts which of those will actually be immunogenic, ie, displayed on tumor cell surface and trigger a T-cell response. It's designed to keep learning from accumulated clinical and immunogenicity data across patients, not just per-patient.
INTerpath-001 (Stage IIB-IV resected melanoma, 2:1 randomized): combination with pembrolizumab beat Keytruda alone on both primary (RFS) and key secondary (DMFS) at interim. Phase 2b at ASCO 2026 showed 49% reduction in recurrence/death, 59% in distant metastasis/death at 5 years. Phase 3 confirmed both.
What this validates:
— tumor-specific neoantigen prediction by ML works in a blinded trial at scale
— 8-week personalized mRNA manufacturing is operationally real
— effect is additive on PD-1 blockade, not redundant
This is first positive Ph3 for individualized neoantigen therapy. First positive Ph3 for any mRNA cancer therapeutic.
What a great time to live in!! This is the best time for AI & biotech!
Moderna and @Merck today announced positive topline results from the Phase 3 INTerpath-001 trial evaluating adjuvant treatment with intismeran autogene, a novel investigational mRNA-based individualized neoantigen therapy jointly developed by Merck and Moderna, in combination with KEYTRUDA® (pembrolizumab), Merck's anti-PD-1 therapy, in patients with completely resected Stage IIB-IV melanoma.
Read more:
octave van den Bosch retweeted
Last week the hydrogen electrolyzer team @terraformindies pulled off yet another first, the sustained production of >99.9% pure H2 from our vertically integrated, California-manufactured electrolyzer stack while it was coupled directly to a solar array at our Muroc desert test site.
WTAF?
Most commercial electrolyzers need carefully conditioned power from expensive battery-meditated backup systems. Ours runs directly off the sun. Clouds pass, the day turns to night, and we maintain purity from a stack whose Bill of Materials cost is well below $100/kW.
Terraform's single-minded focus on capex reduction has allowed us to convert sunlight to hydrogen in an unprecedented demo at a cost below $2/kg. If that wasn't enough, we have a crystal clear plan of steady execution to push that cost below $1/kg in the coming years.
Rather than linger on this point in response to experts-with-spreadsheets who said this was not only beyond my team, it was forbidden by known laws of physics, let me tell you a bit about how we actually pulled this off.
We started building this test site in March. Once the panels and electrics were in place the CO2 team were the first to demonstrate production on site. Close behind them the electrolyzer team planned the logistics necessary to project substantial operational ability into a hostile test site in the middle of nowhere. It's no good to find you're missing a wrench half way through the day!
Much of the test prep was completed before dawn, when the panels go live. The sun came up and the stack immediately started splitting water into hydrogen and oxygen. The team carefully monitored purity and flammability as the sun climbed through the sky. As designed, the stack warmed up and conducted even more power, maintaining solid production until late afternoon when the setting sun shaded the panels.
Terraform's synthetic fuel system is uniquely designed to follow the sun and extract the maximum possible value from cheap solar panels.
Hydrogen is a pernicious molecule. It leaks through and embrittles metals, burns almost invisibly at a wide range of mixtures in air, burns hot and fast and can easily undergo detonation transition, and has about half a dozen other spookily dangerous properties. My advice is to never work with it unless you absolutely have to. The Terraformer produces and consumes H2 in one compact discrete area with a minimum of complexity and fuss, and as expected this demo was completed in accordance with our rigorous safety standards and no unscheduled excitement!
This successful demonstration was also a profound milestone for the team after a testing anomaly last December compelled us to finally rip off the bandaid and move decisively towards the "future design" with half the parts but considerable complexity in assembly. No-one else makes electrolyzers this way and we, more than anyone, know exactly why. And also how to do it anyway, translating directly into a unique cost advantage.
A huge congratulation to Ken, Sherman, @ckalitin, Nikhil, Abdullah, and Aaron for their successful test campaign.
Terraform's hydrogen and CO2 are the chemical precursors for synthetic methane and methanol, which we make in our own synthetic fuel reactor.
Combined, we make oil and gas out of sunlight and air. We are breaking the geological and geographical monopolies on oil production.
In the limit, Terraform will deploy these electrolyzers by the millions and they will all be plug-and-play with solar PV arrays.
The Muroc smoke test campaign is far from over.
We will win!
octave van den Bosch retweeted
JUST IN: Investor Marc Andreessen says "Doctor ChatGPT is a better doctor than 99% of doctors".
octave van den Bosch retweeted
“The limits of my language are the limits of my world.”
—Ludwig Wittgenstein
octave van den Bosch retweeted
My advice to founders in 2026: spend tokens, not headcount.
Record everything. Make your company queryable. Build self-improving loops.
Before long, AI won’t just help you operate your company. It will make it self improving.
Don't think AI adoption, think AI transformation.
This is the biggest shift in how startups get built since cloud computing.
octave van den Bosch retweeted
Three days ago I left autoresearch tuning nanochat for ~2 days on depth=12 model. It found ~20 changes that improved the validation loss. I tested these changes yesterday and all of them were additive and transferred to larger (depth=24) models. Stacking up all of these changes, today I measured that the leaderboard's "Time to GPT-2" drops from 2.02 hours to 1.80 hours (~11% improvement), this will be the new leaderboard entry. So yes, these are real improvements and they make an actual difference. I am mildly surprised that my very first naive attempt already worked this well on top of what I thought was already a fairly manually well-tuned project.
This is a first for me because I am very used to doing the iterative optimization of neural network training manually. You come up with ideas, you implement them, you check if they work (better validation loss), you come up with new ideas based on that, you read some papers for inspiration, etc etc. This is the bread and butter of what I do daily for 2 decades. Seeing the agent do this entire workflow end-to-end and all by itself as it worked through approx. 700 changes autonomously is wild. It really looked at the sequence of results of experiments and used that to plan the next ones. It's not novel, ground-breaking "research" (yet), but all the adjustments are "real", I didn't find them manually previously, and they stack up and actually improved nanochat. Among the bigger things e.g.:
- It noticed an oversight that my parameterless QKnorm didn't have a scaler multiplier attached, so my attention was too diffuse. The agent found multipliers to sharpen it, pointing to future work.
- It found that the Value Embeddings really like regularization and I wasn't applying any (oops).
- It found that my banded attention was too conservative (i forgot to tune it).
- It found that AdamW betas were all messed up.
- It tuned the weight decay schedule.
- It tuned the network initialization.
This is on top of all the tuning I've already done over a good amount of time. The exact commit is here, from this "round 1" of autoresearch. I am going to kick off "round 2", and in parallel I am looking at how multiple agents can collaborate to unlock parallelism.
github.com/karpathy/nanochat…
All LLM frontier labs will do this. It's the final boss battle. It's a lot more complex at scale of course - you don't just have a single train. py file to tune. But doing it is "just engineering" and it's going to work. You spin up a swarm of agents, you have them collaborate to tune smaller models, you promote the most promising ideas to increasingly larger scales, and humans (optionally) contribute on the edges.
And more generally, *any* metric you care about that is reasonably efficient to evaluate (or that has more efficient proxy metrics such as training a smaller network) can be autoresearched by an agent swarm. It's worth thinking about whether your problem falls into this bucket too.
octave van den Bosch retweeted
ICON has unveiled a next-gen construction 3D printer that replaces older gantry systems.
Moving on tracks between sites, it cuts setup time and labor.
With a 70-ft print radius, it can build multi-story structures up to 27 ft tall and 110 ft long faster and cheaper.
octave van den Bosch retweeted
Meta to cut 10% of jobs, or 8,000 employees, report says techcrunch.com/2026/04/23/me…
octave van den Bosch retweeted
The first simulated image of a black hole, calculated with an IBM 7040 computer using 1960 punch cards and hand-plotted by French astrophysicist Jean-Pierre Luminet in 1978.
