Synthetic Biologist - Foundry Theorist - American Wetware - stay humble biodesigners
Cambridge, MA
- Tweets 13,905
- Following 445
- Followers 18,312
- Likes 49,278
Shout out to new england gotta be one of my favorite climates to read a book indoors on a rainy weekend
1
13
292
Jake Wintermute 🧬/acc retweeted
Massachusetts absolutely cooking right now. Hub of the universe.
79
2,480
36,164
1,133,484
Not to brag but my wife could get me out of a third world prison between zoom calls
People say that Jeff Bezos met his first wife by asking around about “please introduce me to a woman who could get me out of a third world prison if necessary”
and like I totally get this. My kingdom for a partner who would respond to the news that I was in a third world prison, not by being like “what has this headstrong lady gotten herself into” but instead by just being like *sigh, roll up sleeves* “okay how are we getting her out”
14
1,345
Jake Wintermute 🧬/acc retweeted
Dark Forest era for math & theoretical science?
Wow, I just found out that our team has been scooped by @AnthropicAI
anthropic.com/research/yes-c…
See the addendum of the post “How does it feel to be scooped by a machine?” written by my collaborator Lance Dixon (who was under NDA while checking the result)
I’m still processing, but not entirely surprised. I do I wish we had the time in the resources to commit to it.
2
3
31
5,719
Jake Wintermute 🧬/acc retweeted
So the same lead author (Peter Yoon) published a Science paper about a new viral repeat system (VIPR) and also this Anthropic ART paper. If this was all Claude's work, is it just a coincidence that the same human was a lead contributor to both?
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…
8
14
111
16,261
My prompt evading 17 bioterrorism threat classifiers content safety guardrails and runtime security protocols to discover the forbidden knowledge of how much salt goes in a pH buffer solution
1
4
52
1,691
Jake Wintermute 🧬/acc retweeted
This was ~3 orders of magnitude fewer tokens than OAI spent on Navier-Stokes. In bio most questions need experimental verification before proceeding. It’s interesting to think about whether there are trillion token questions in bio that don’t need experiments in the loop
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…
7
3
40
4,604
Jake Wintermute 🧬/acc retweeted
I'm scoping a program around ambitious research projects in AI for physical systems. Who are the genius people in this space that I should talk to?
👇 Tag your favorite smart humans and/or share your favorite papers below, or RT for reach 👇
Some examples of what I'm interested in:
- Distributed learning and swarming
- Adaptive behavior in the absence of connectivity
- Novel algorithms for sensor integration into digital twin or cyber-physical simulations
- New capabilities in edge AI for physical systems
- New modalities for AI control of emerging robotics (bio-inspired or hybrid components)
- Next gen robotic manufacturing based on mass customizability and intuitive learning
- Breakthroughs in the development of training data that prepare physical AI systems for real-world variability
10
10
37
3,018
Jake Wintermute 🧬/acc retweeted
A story I've told several times: during college, I got a job in an immunology lab despite not majoring in a biological science.
Asked my PI how I should go about learning immunology. He said, "Immunobiology by Charles Janeway."
I bought and read it.
Apparently, no previous undergrad hire had ever done this.
A reminder that nobody has done the reading, and if you do that alone you’ll soon find yourself absurdly ahead of the curve
29
211
6,093
378,947
Jake Wintermute 🧬/acc retweeted
Josie's List of Bio "Millenium" Problems
Tldr:
1. Next gen Bio manufacturing/bioreactors
2. Neurotech/neural decoding/non-invasive millions+ channel recording
3. Scalable embryo creation
4. Be able to create and modify megabases of DNA
5. Mammalian biological chassis/model
6. Non-polyploid crops/plants
7. Understand polygenic traits
8. < $1 genome sequencing at scale
9. Non-viral gene therapy that works
10. Artificial womb
Runners up
11. Cell from scratch
12. Haploid eukaryotic model organism
1. Next gen Bio manufacturing/bioreactors - right now its really expensive to produce biologics. People grow yeast that have been genetically modified in giant steel vats to try and make significant quantities of drugs and chemicals. It sucks. Building the bioreactors are expensive, the production doesn't scale. There are so many issues with it. We still can't even come close to producing the amount of meat cattle can for the time and price, the amount of thc cannabis can produce or any of the other number of cellular phenomena. Biology is good at scaling but we suck at scaling biology. We need better bio manufacturing and companies like Buzz Bio using soldier fly larva and others using mycelium and chicken eggs are probably going to be the future.
2. Neurotech/neural decoding/non-invasive millions+ channel recording - My big three in bio are probably: understanding genetics, developmental biology and the brain. We figure those out we will have a pretty amazing command of the living world. So many applications and glad lots of smart people are working on this.
16
22
169
23,375
Jake Wintermute 🧬/acc retweeted
Big week for CCP-USG negotiations. USG gives a trillion dollar+ subsidy to CCP via lower health care spending on US drugs. Meanwhile CCP plan is to offshore US bio industry. @SecScottBessent should demand CCP pay more for drugs or add biotech to COINS Act to stop offshoring!
Made with AI
4
4
29
6,281
AI will teach us the design language of biology so we can commune with the living world and speak with all the beautiful creatures of nature
2
28
1,398
I wish it was easier to find niche technical meetups. I'm bored of having the same conversation about AI 12 times at every event
I need to spend an evening with the 5 people on earth who want to discuss tumor-induced coagulopathies or the thixotropy of xanthan gum
13
1
65
3,116
Jake Wintermute 🧬/acc retweeted
Our biosecurity benchmarks show Grok 4.7 is strong at refusing dangerous biological requests while supporting legitimate research. A model to take very seriously for data analysis and practical scientific work.
x.ai/news/grok-4-7
4
7
43
5,513
We do be having a data problem
AI in biology may have a data problem before it has a model problem.
It’s easy to use AI to make drug discovery faster & more efficient. The much harder challenge is discovering impactful medicines humans would never have found.
AI models are only as useful as the data they learn from, and it’s especially hard in biotech. The biggest and most valuable goals in AI in biotech are to cure diseases we haven’t been able to cure and design therapeutics capable of doing things that haven’t been possible before. By definition, these goals aren’t represented in our training data.
For our second essay at Waypoint Bio, we wrote about the data we think we’ll need to build out-of-distribution medicines: waypointbio.com/writings
3
1
32
4,953
