@buckfiftyseven@mastodon.social
Post #4214826
2026-07-28 17:22 UTC
As an aside, I do know someone who does cancer research at a major cancer center. They are training an internal AI, and testing it against their processes.
They bring it all in house for a few reasons of course. The first is that they can protect patient privacy. The second is that they can develop their own methods. The third is that they can test it according to their own standards.
Replies (1)
-
@ApostateEnglishman@mastodon.world 2026-07-28 19:02
@buckfiftyseven@mastodon.social It's the "AI" evangelists who are like antivaxxers following a zombie cult; those opposed to LLMs and similar systems are mostly rational, and doing so based on facts and evidence. There's a mountain of reports and studies to draw from. You have it backwards. There's nothing wrong with specialised machine learning trained on experimental data for use in well-defined scientific contexts, such as developing new medicines (your example). That's not what anyone here is taking about.