In partnership with
News of the day
1. OpenAI Foundation is funding the creation of high-quality scientific datasets, including failed biotech company archives, to advance AI in medicine and drug discovery. → Read more
2. Reward AI unveils OM-1, a novel robot policy trained solely on human demonstrations, bypassing traditional teleoperation and on-robot data for versatile manipulation. → Read more
3. Chinese researchers propose a five-level framework to measure AI self-improvement, highlighting current limitations and future challenges. → Read more
4. Agent-net releases Webagent, an open-source Go harness enabling businesses to transform websites into guarded AI agents with a declarative JSON spec. → Read more
Our take
Hi Dotikers!
AI in medicine has a dirty little secret: the models are hungry, and the pantry is nearly empty. This week's story tackles that problem head on. The OpenAI Foundation, the nonprofit arm sitting on a stake in OpenAI that could soon be worth 250 billion dollars, just launched Data for Public Health, a grant program designed to fund the creation of high-quality scientific datasets for biology and medicine.
The headline numbers are impressive: 40 million dollars for cancer vaccine data at the University of North Carolina, support for OpenAdmet's drug prediction competitions, and a target of 1 billion dollars in grants by year's end. But the most fascinating idea is also the cheapest. Analyst Ruxandra Teslo received 500,000 dollars to pursue what she calls biotech's lost archive: buying the regulatory filings, safety data and manufacturing secrets of bankrupt biotech companies at auction. Turns out one company's funeral is another algorithm's breakfast.
This is genuinely smart. Failed drug programs contain enormous scientific value that usually vanishes into legal limbo, and recycling it into training data could make drug development less of a black box for everyone. It deserves applause. What deserves scrutiny is the emerging pattern behind it. Google recently won the corporate data of Spirit Airlines in bankruptcy, including 100 million emails, sparking real privacy concerns. Bankruptcy courts risk becoming the new frontier of the data land grab, where the interests of patients, employees and trial volunteers get settled by whoever bids highest. The medical case here is far more defensible than hoovering up airline inboxes, but the mechanism is the same, and it needs rules before it needs momentum.
If AI is going to cure anything, it will need more data. The question this week poses is simple: who decides how we get it, and at what price?
Alex.
Some teams never seem to stop moving. They're on Attio, the agentic CRM.
Every customer signal is captured in one shared context layer, always current and compounding. Agents and workflows build pipeline, chase every buying signal, and move deals forward, an always-on revenue engine running alongside your team.
With Attio, you’ll get:
Leads automatically prioritised and routed to the right rep
Expansion and risk signals caught the moment they land
Follow-ups written in your voice, already there when you arrive
Teams like Parallel, Turbopuffer, and Wordsmith build on Attio. Are you one of them?
Meme of the day
