Either it needs to run on our hardware, or we need explicit contractual guardrails around use of the data for their purposes. Obviously anything you give Anthropic etc. is going into their training pipeline for instance, and they change their terms of use every week so it's impossible to work with as a small company with no leverage
Our take has been that if you are working with one of these large companies, you cannot expect anything to be private. We've worked with a smaller more focused product where we were able to negotiate the necessary controls
There is some history to why it is as grim as it is right now (sunny skies will return, don't worry!). A lot of funding was around ~2019 from the VC people, and biomed PIs were getting their startups funded and hiring from the recent PhD cohorts. The horrible environment in STEM academic hiring needs no introduction, so the talent pool was rich. Early stage drug development and biotech is horrendously risky, so most don't make it past ~5-7 years. Now there are lots of people looking for a job, and the only surviving companies are the extremely hiring averse ones.
Drug development is just a totally different game. The tools are the same but the difference between what the reviewers at your favorite high impact publication want and what the FDA wants are pretty different. People spend their whole careers getting good at the latter in the same way people get good at the former.
I've seen people come from academia and thrive and I've also seen the struggle. Some people also go to school with the goal of doing drug development, which sometimes academic folks don't realize.
- person who was good at microscope and ended up in early stage drug development ~10y