Our Kaggle competition to predict small molecule binding is over, and nobody was able to generalize into novel chemical space whatsoever. There's a lot of work to do! Blog post: lnkd.in/gW7VNaEa Endpoints: lnkd.in/gUefHgJg

A short summary: axios.com/2024/06/23/leo… Leopold Aschenbrenner — formerly of OpenAI's Superalignment team, now founder of an investment firm focused on artificial general intelligence (AGI) — has posted a massive, provocative essay putting a long lens on AI's future. Why it

Scalable in vivo experiments provide the most complete biological context for therapeutics discovery. This is why NOETIK we couple human data with a high throughout in vivo perturbation platform (Perturb-Map).

Tony Wood on AI for molecule design: ...I would say that’s in many ways polishing diamonds. We’re pretty good at that anyway. And if I solve that problem, I’m not solving the real problem, which is that most things fail in Phase 2, going back to targets and patients. 💯

Our Kaggle competition got a nice writeup by Derek Lowe. It's true, predicting novel chemical material is hard! science.org/content/blog-p…

Enjoyed discussing our $40M Series A financing led by Polaris Partners with Andrew Dunn!

We would kick off the work days at early Recursion with a 6am summit of local Mt Wire. With Chris Gibson & Mason Victors catching up and testing out these sea level lungs again this morning. Can still do it (barely). LFG. 🚀

ftw 😂

I get asked a lot for startup advice. You’re failing, what’s the secret, etc My 2 cents: > There’s a GREAT FILTER in startups that kills almost everyone. It’s there to weed out the ones who are not meant to be there > But it can be beaten, even with little or no capital. If

Excited to be partnered with Breakout Ventures for NOETIK series A! Check out their blog to read more about our mission.

Excited for NOETIK to participate in the Amazon Web Services GenAI Accelerator. We’re training first-in-class frontier models that understand human cancer biology, and looking forward to scaling! businesswire.com/news/home/2024…

I’m thrilled to announce: at JURA Bio, Inc. we’ve constructed a new kind of generative protein model, which allows us to manufacture its generated designs in the real world at petascale. Paper: static1.squarespace.com/static/66db3a5… Blog: jura.bio/blog/variation…

Folks learning for second time around that without differentiated data, models get commoditized quickly. Faster this cycle even.

Neoadjuvant immunotherapy in locally advanced MMR-deficient colon cancer: #ESMO24 🔎3-year disease-free survival from NICHE-2 👉💯% DFS 👉all MRD pts cleared ctDNA 🧐Impressive results, ctDNA may be helpful to select for organ preservation ESMO - Eur. Oncology #ESMOAmbassadors

This is why we’re focused on cancer biology rather than pharmacology.

Amazing NOETIK team offsite this week in Park City, UT. Good mix of intense productivity… and equally intense connectivity.

To solve hard problems in bio x ai, we need to start from the right data. Check out our recent post Tackling Cancer as a Data Problem on NOETIK noetik.ai/lungcanceratlas