Building something new for human enhancement
San Francisco, CA
Joined September 2012
- Tweets4.8K
- Following2.8K
- Followers1.1K
- Likes12K
David retweeted
This is why we call it the “Pilot In Command”
This happened in the same vicinity where the ill-fated Lear Jet hit the storm cell.
JetBlue pilot en route from Fort Lauderdale to Austin says to ATC:
“So you're gonna make us go through a, uh, severe thunderstorm?”
The captain then commanded the situation and overrode the ATC:
“So we declare an emergency, making a left turn heading 260 for severe weather avoidance.”
Sometimes piloting demands a specific set of traits - decisiveness, independent judgment, and the backbone to reject an instruction when it doesn’t feel right.
Refusing the assigned heading showed exactly that kind of assertive leadership.
Whoever is pilot monitoring must make these tough calls but ultimately, the captain is the final authority and ultimately accountable for every soul on board.
This needs to be said more often. I'm all for progress, but we should always remember the many people across many organizations that make progress possible
The other piece that often gets missed in the recent discussion of lean biotechs post-Mithril announcement is that there are always a large number of FTE equivalents within CROs working on any virtual biotech.
There’s a legitimate corporate structure decision to outsource vs insource, but no biotech can function without a lot of people working at the bench, whether or not they’re on payroll.
(Yes I’m aware of the concept of the autonomous lab, but so far there are none for in vivo studies or many other essential tasks in biotech).
Ooo i had never considered this.
The Fruits of AI Drug Abundance: Indigestion In The Near Term
While the long term impact of AI is likely transformational, in the near term more drugs against known targets will likely exacerbate the crowding challenge dragging down industry R&D productivity
lifescivc.com/2026/10/the-fr…
This was a fascinating read. Wow.
The Structural History of Eukarya (SHE): The first proteome-scale phylogeny constructed entirely from 3D structure. We computed 300 trillion alignments across 1,542 species to map the tree of life. Now published!!! science.org/doi/10.1126/scia…
"but we're stopping short of clinical trials". Suuure.
One thing people missed is that @AnthropicAI has experienced drug development veterans running their drug discovery effort.
Michael Pourdehnad is former SVP of BMS Early Clinical Dev. for Hematology and Oncology.
David retweeted
I need all the AI people to stop saying "biology" when they mean "medicine" or "drug development"
David retweeted
🥈Second place in this year's #NikonSmallWorld photo competition goes to John-Oliver Dum for his image of a jumping spider (Salticus scenicus) with pollen. Cute but spooky, and just in time for Halloween next weekend! 🕷️🎃
bit.ly/4hYjUtw
ALT Close-up image of a jumping spider covered in pollen, showing its large eyes and detailed hair-like structures.
My biggest concern today is assimilation to the mean. So many people are discussing exactly the same ideas and ignoring so many fantastic approaches.
Very cool! People should sign up and contribute to building resources like this
Superintelligence should benefit human intelligence. Mathematics is our starting point.
We’re launching Readable Intelligence, and we invite the math and AI communities to build it with us.
📄 Pick a paper.
💬 Flag an unclear step.
🙌 Contribute a more readable version....
readable-intelligence.com/
David retweeted
The ideal outcome for science isn't a couple of labs using proprietary models and millions of dollars to crack trophy problems, it's to have millions of AI-aided human scientists who follow their curiosity and become way more ambitious
This looks awesome.
1/ Most genes and variants matter somewhere but look silent in any one cell model. Can we build conditions that reveal phenotypes? With @kanishkadey, we engineer cell states to expose hidden disease biology and trace how genetic risk reaches disease genes. biorxiv.org/content/10.64898…
ALT Visual abstract: activating TFs singly and in pairs engineers new cell states, and the Mosaic model learns how TFs combine. Findings: hidden AMD risk variants, state-dependent convergence, and chromatin priming.
This already happens. We don't know the MOAs for most drugs. Pharma *does* game Ph3 trials (larger trials are better powered, but also higher likelihood of adverse effects, optimal trial size arrived at thru cost optimization). Safety does *not* require mechanism.
I would happily bet anything that this newest batch of problems means OpenAI is pursuing fusion and quantum computing solutions--the joint answers to 'how do we get enough energy to make AI work'.
Thoughtful take by a major figure in Dev bio. I would add, however, that we should also be pushing biologists to engage with AI and AI research, rather than passively waiting for AI researchers to engage with biology. Our community makes this mistake often
Recently, preparing a talk made me think @ the impact of AI in #DevBio and led me to ask AI what would it do to explain the SM #organizer expt just w/ data w/o conceptual input from human interpretation of experiments. Its answer below 🧵
