zhaoziyuan2000

Would you mind to clarify what you mean by standardized multi-modal ground truth?

There are certainly non engineering problems out there... For one example, in biomedicine it's known cancer in different organs can have preference of having immune hot (in skin and lung cancer, e.g.) and cold tumors (in pancreatic and prostate cancer, e.g.), and current immunotherapy work much less well on the latter. Why this pattern exists is still an open question very important for treating patients and it's not a how did nature engineer question as it's a unintended byproduct from several tiers of complex biology involving understanding mutational burden in tumor cells, immune surveillance surrounding the tumors, and mechanical structure of the tissue, etc. The engineering payoff comes later, i.e., immunotherapy works better on a particular subset of patients, after explaining and understanding the biological phenomenon first.

It seems that in a frantic pace to prove general usefulness of these LLMs the AI companies are also picking up the role of traditional scientific researchers without really have enough proper communication with the scientific community at large... I agree putting tools in the actual scientists will be very helpful but do they have the patience waiting or even parsing their feedback?

Does this remind anyone of that novel Flowers for Algernon..?

This is really cool! Also, my girlfriend suggested it would be nice if she can login and be on the same flight so we can fly home together.