Malte Kuehl

NYT headline: First 6 days of Iran War Cost US $11.3 billion, Pentagon says

The ERC budget for 2026 is 2.7 billion euros 🧪

Excited to share this preprint that describes my latest work on using GPUs to accelerate processing of RNA-seq data.

The title says it all: "RNA-seq analysis in seconds using GPUs" now on biorxiv www.biorxiv.org/content/10.6... and github github.com/pachterlab/k...

Figure 1 shows they key result

having used claude extensively for the past couple of months, i think there is still a lot of value to understanding your code and steering it intentionally. the amount of incidental complexity claude accumulates otherwise (which mostly results in bugs that it can't ever fix reliably) is staggering.

Es zieht ein neuer Ton ein in die Redaktionen. Die Bombardierung von Zivilist*innen ist «Drecksarbeit», das Völkerrecht nicht mehr als «Gerede». Doch ist man keinesfalls gefühllos und ohne Empathie. Wenn Faschisten öffentlich weinen, ist man ganz gerührt …

This sounds like an ad, but it's about murder.

Anybody who has used these models should be deeply concerned about this.

Der Bundeskanzler sitzt neben Trump, während der Spanien droht und lästert, und stimmt eifrig ein. Dass Autokraten hofieren nicht funktioniert, wird gleich gezeigt - Trump droht mit Zöllen gegen Deutschland, lacht bräsig und schlägt Merz aufs Hosenbein. Vorgeführt und blamiert. Eine Schande.

In Zeiten von Budgetkürzungen und Angriffen von Rechts auf die freie Wissenschaft ist der letzte Absatz ein absoluter Offenbarungseid.

"Erkennt die Bundesregierung an, dass die USA und Israel einen Angriffskrieg führen?"
"Nein, das erkenne ich nicht an"
- Regierungssprecher Kornelius

Meanwhile Germany as evidenced by the E3 statement has chosen to become party to a foreign war and German media is not even talking about it… Shameful!

‘Unbelievably dangerous’: experts sound alarm after ChatGPT Health fails to recognise medical emergencies Study finds ChatGPT Health did not recommend a hospital visit when medically necessary in more than half of cases

Study finds ChatGPT Health did not recommend a hospital visit when medically necessary in more than half of cases. www.theguardian.com/technology/2...

I apologize for initially being somewhat skeptical of Phillip Morris.

In all of our interactions, they made it profoundly clear that they are unequivocally opposed to any and all use of tobacco products and that they remain committed to partnering with us to achieve the best possible outcome.

And the Lancet's editorial that goes with it
www.thelancet.com/journals/lan...

Agentic AI and the rise of in silico team science in biomedical research - Nature Biotechnology This study examines the emerging use of agentic artificial intelligence (AI) in biomedical research, highlighting the challenges and opportunities that may inform the design of agentic AI systems suit...

Our perspectives piece on Agentic AI for biomedical research has been published in Nature Biotechnology www.nature.com/articles/s41... #agenticAI #artificialintelligence #agents

The political effects of X's feed algorithm https://doi.org/10.1038/s41586-026-10098-2 Received: 16 December 2024 Accepted: 4 January 2026 Published online: 18 February 2026 Open access • Check for updates Germain Gauthier,5, Roland Hodler?5, Philine Widmer35 & Ekaterina Zhuravskaya3,4,5 m Feed algorithms are widely suspected to influence political attitudes. However, previous evidence from switching off the algorithm on Meta platforms found no political effects'. Here we present results from a 2023 field experiment on Elon Musk's platform X shedding light on this puzzle. We assigned active US-based users randomly to either an algorithmic or a chronological feed for 7 weeks, measuring political attitudes and online behaviour. Switching from a chronological to an algorithmic feed increased engagement and shifted political opinion towards more conservative positions, particularly regarding policy priorities, perceptions of criminal investigations into Donald Trump and views on the war in Ukraine. In contrast, switching from the algorithmic to the chronological feed had no comparable effects. Neither switching the algorithm on nor switching it off significantly affected affective polarization or self-reported partisanship. To investigate the mechanism, we analysed users' feed content and behaviour. We found that the algorithm promotes conservative content and demotes posts by traditional media. Exposure to algorithmic content leads users to follow conservative political activist accounts, which they continue to follow even after switching off the algorithm, helping explain the asymmetry in effects. These results suggest that initial exposure to X's algorithm has persistent effects on users' current political attitudes and account-following behaviour, even in the absence of a detectable effect on partisanship.

A new paper shows that less than 2 months of exposure to Twitter’s algorithmic feed significantly shifts people’s political views to the right.

Moving from chronological feed to the algorithmic feed also increases engagement.

This is one of the most concerning papers I’ve read in awhile.

Satellite proposals threaten the night sky In the United States, the Federal Communications Commission (FCC), the agency responsible for authorizing satellite launches and operations…

The FCC just opened public comments on SpaceX's plan to launch a million satellites to do AI compute in space. Under the current proposal, an environmental review won't be required. Please consider submitting a public comment to oppose this damaging plan.
darksky.org/news/two-sat...

“No one’s going to say to students after they graduate, ‘Do the best job you can, but if you use AI, you’ll be in trouble,’” Roberts told me. “Yet we have faculty members effectively saying that right now.”

I also loathe this line. Enforcing *learning* concepts in the classroom is not the same thing as what is expected in *applying* those concepts in the workplace. AI is a powerful tool for the prepared mind, it's our job to prepare those minds.

"Moderna's CEO announced the company will no longer invest in new Phase 3 vaccine trials for infectious diseases: 'You cannot make a return on investment if you don't have access to the U.S. market. Vaccines for Epstein-Barr virus, herpes, and shingles have been shelved.” 9:12 AM Feb 12, 2026 604.8K Views

In terms of what cures are being lost:

- Epstein-Barr virus is perhaps the major trigger for multiple sclerosis
- herpes simplex virus causes cold sores, genital herpes, infections in babies, deadly meningitis
- shingles virus causes an intensely painful disease

Beyond the sheer absurdity of the situation. This is harmful and deeply antisocial behavior by the person who created this agent, set it loose and continues to direct its behavior.

Bad enough to spam open source maintainers, but hijacking the language and plight of actually marginalized humans.

A bar chart showing the electricity use of several daily activities with the subtitle "The 'typical query' is not a useful way to think about coding agents' energy use." The bar for a 'typical ChatGPT query' is not even visible. My median Claude Code session is somewhere between the average US household per minute and toasting bread for three minutes. My median day with Claude Code is something like running a dishwasher.

Whenever I read discourse on AI energy/water use that focuses on the "median query," I can't help but feel misled. Coding agents like Claude Code send hundreds of longer-than-median queries every session, and I run dozens of sessions a day.

On my blog: www.simonpcouch.com/blog/2026-01...

Anoctamin-2-specific T cells link Epstein-Barr virus to multiple sclerosis Researchers identified anoctamin-2 (ANO2) as a frequent autoimmune target in multiple sclerosis, with T cell responses against ANO2 occurring in over half of patients. These ANO2-specific T cells shar...

Epstein-Barr Virus #EBV was linked to #MultipleSclerosis - now a plausible cause has been found, misidentification by longterm memory T-cells that pick on the wrong protein, ANO2, instead of the EBV-antigen. Massive inflection towards the elimination of MS!!!
🧪🧠Cell www.cell.com/cell/fulltex...

If Google truly wants to increase its API market share, trust needs to be earned.

@developers.google.com Benchmark performance is one thing, but reliable systems that devs can rely on are another. Gemini 3 API has been broken for ~14 hours now (responding in JSON, not natural text), with dozens of reports across Gemini forum, GitHub... Yet no solution. Please escalate internally.

Proud to close out an incredible year with an update on where scverse is today and where we’re heading.
I’d love for you to join me, share your ideas, challenge us, and help shape what comes next. It’s time! 💪

Connecting AI to biology: Model Context Protocol Julio Saez-Rodriguez discusses how MCP servers are helping AI connect to life science databases, making AI a more reproducible tool for research.

AI holds huge promise for life science research. But unreliable answers currently limit its impact.

@juliosaezrod.bsky.social, Head of Research at EMBL-EBI, discusses how Model Context Protocol (MCP) servers are helping to change this.

www.ebi.ac.uk/about/news/p...
🧬💻

What a suitable moment to have a community meeting by @maltekuehl.com and @biocontext.ai! Since we didn’t manage to have him on the conference, join us next Tuesday! 💪❤️

We will have our next community meeting on Tuesday, 2025-11-25 at 11:00 CST!

Malte Kuehl (@maltekuehl.com) will be presenting on BioContextAI!

More information and Zoom registration link in 🧵

🤖🧬 New biomedical MCP server added to the registry:
Nucleotide Archive MCP server
Check it out at https://biocontext.ai/registry/biocontext-ai/nucleotide_archive_mcp

🤖🧬 New biomedical MCP server added to the registry:
anndata-mcp
Check it out at https://biocontext.ai/registry/biocontext-ai/anndata-mcp

The protein concentration in the cytoplasm is so high that the average protein has a water hydration shell with a thickness of only ≈ 10 water molecules separating it from the adjacent protein hydration shell.

1/2