Control strategy for large scale bio datasets serving as substrate for AI is its own science (and one I’d suggest we played a very meaningful role in pioneering). Can’t agree more with Ron. If you see controls in a row, column or on the edge of your plate, rather than randomized across a plate, you know your dataset is not built with ML/AI in mind.

A lot of wisdom from a person and team @_DimensionCap that has been at the bleeding edge of Tech x Bio longer than almost anyone. The @CoefficientBio x @AnthropicAI is just the beginning from this team.

Check out my latest article: Fifteen Years. One Conversation. linkedin.com/pulse/fifteen-y… via @LinkedIn

Would you rather own two TBPNs or one Coefficient Bio? theinformation.com/articles/…

News! @airstreet has raised $232,323,232 for Fund III to back AI-first companies from the earliest stages in the US and Europe. Now the largest solo GP venture firm in Europe. Our third epoch begins today. Join us!

Heck yes!

Freed from the shackles of daily operation, and with a few weeks of decent sleep under my belt, I found the time to sit down and write a thing about TechBio. Take a read and then tell me if you are an 'AI maximalist' or an 'AI minimalist'... (yes, that's a leading question). linkedin.com/pulse/wrong-ben… via @LinkedIn

Joining the @PacBio board. In the AI era, data quality is the moat. Doesn’t matter how good your model is if your training data is noise. I believe PacBio makes the best sequencing data on the planet. Time to help them turn that into a category-defining advantage. 🧬 investor.pacificbiosciences.…

Seems fine…

Autonomous delivery still rusty…

Fantastic work! So excited to watch this vision grow wings! 👏👏👏

Love this.

Adding this to a growing list of signals that suggest not that we are necessarily at some tipping point in the world today, but that we seem to be accelerating in that direction in a way that feels less and less recoverable except by extraordinary means. It seems to me AI isn’t the driver of our trajectory, but simply an accelerant. After all, these tools are trained on the collective output us humans.

Go hawks!

🚀 How Recursion used AI to identify a novel approach for potentially treating the rare disease Familial Adenomatous Polyposis (FAP). A new video shares the story of Recursion’s investigational drug REC-4881 for the rare disease FAP which affects more than 50,000 people in the US and EU5 and has no approved medicines. Patients with this disease develop hundreds to thousands of polyps in their early teens to 20s that, if left untreated, have a nearly 100% risk of developing into colon cancer. ⬇️ Here’s how we did it: ▪️ Using the Recursion OS platform to analyze cell morphology, we tested thousands of compounds to find drugs that restored FAP-diseased cells to a healthy state. ▪️ This led us to REC-4881 — a drug that blocks a biological pathway called MEK1/2 – a new approach for treating FAP that had not been investigated clinically before. ▪️ Now, we’re using the full Recursion OS 2.0 platform — including our ClinTech real-world evidence –  to fully understand disease burden and guide our clinical strategy, allowing us to expand patient eligibility for our trial. 💡 The takeaway: This program is the first clinical validation of the Recursion OS, and demonstrates our ability to translate unbiased phenotypic insights into potentially differentiated treatments for diseases with high unmet need. 👉 Read the full summary data in our December 8, 2025 press release here: ir.recursion.com/news-releas…

“When the institutions are optimized to fund the legible thing and the individuals are optimized to build the legible thing, the identity of the founder itself fades. Being a founder used to carry with it the implication that you had seen something others hadn’t, that you were willing to be wrong in public about an unlikely future you believed in, and take a risk of banging your head against a wall with high career risk for a long time.” Definitely seeing this… however, I (hope) and believe there will be a big wave of mission-driven companies, even with millennial founders, as people realize now that the pendulum has swung how thoroughly exhausting it is to build something you don’t feel good about.

Robinhood CEO @vladtenev reinvented their earnings calls after watching a YouTuber livestream one and apologize for how boring it was — then modeled the format after NBA Playoffs postgame interviews: “If I’m bored, the people watching this must be really bored.” “One year ago, we were doing earnings the old fashioned way. We had the Polycom. We had the script. I would read the script. We would take the questions over Polycom. Sell side analysts would be on the call exclusively.” “At one point the livestreamer kind of apologized to the audience. He's like, ‘I know it's a little bit dry. Sorry guys, but we'll get to Q&A soon.’” “What's an example, an analogy, something we could learn from and lift from another industry that is similar to this, but entertaining? I thought, it's kind of similar to an NBA Playoffs postgame interview — you've had something happen… It's either a good quarter, where you're beating and you have good things to say, or maybe it's a bad quarter. But it's kind of entertaining either way, right?” “Q4 of last year was our first earnings call where we did live video… Fast forward to our previous earnings call — it was at the Chase Center. We had live video. We had live Q&A where retail investors would basically livestream and ask their questions live.” “We went from hundreds of listeners on the old Polycom-style earnings to tens of thousands, maybe even 100,000 at the last one. It increased the actual audience of our earnings by multiple orders of magnitude.”

I may be biased, but I’d say the award for most exciting and transformative JPM presentation goes to Najat Khan and ⁦@RecursionPharma⁩ 😉

Huge congrats to the Loyal team. Excited for the future!

Big thanks to @NVIDIAHealth for a fantastic night, kicking off with some Jensen time! Always so energizing to hear his stories and perspective from the last 33 years of building. And humbled to receive a ‘Jensen edition’ DGX Spark in recognition of the pioneering work of the incredible @RecursionPharma team and excited to get that GPU going brrrr! Also extremely exciting to see Dave Rick’s from @EliLillyandCo on stage with Jensen talking about their huge collaboration. While we’ve technically lost the “fastest dedicated supercomputer of any biopharma” title to Lilly for now, their huge investments in the space are a sign that the world is finally understanding the inevitability of TechBio as the future of bio, and that is good not just for Recursion, but for patients and the space. #JPMorgan