Comment by hermitShell
3 days ago
It seems like you're talking about chemical weapons.
Which is fair, they are very easy to make without the assistance of an LLM, and I don't hear about that kind of thing in the news.
I was talking about biological weapons like gain of function and that kind of thing.
The implied question was 'Should companies like Meta publish open weight models when they could theoretically be subverted by malicious users?'. So far as I'm aware there are few people worrying about this kind of thing. Bill Gates for one. But apparently it's off topic and/or not interesting, or not welcomed as a discussion point.
You were worried about someone using LLMs to make weapons. So I went for the ones that are more realistic. Bioweapons are far harder to make than chemical weapons. The science is harder, the actual synthesis can only happen at a few biomed companies (which is where I'd put my checks if I were you), even when you have the DNA synthesis is very hard, in the "just don't bother before your PhD hard" realm, and they're even harder to handle without killing yourself. Bioweapons are not realistic.
To be completely honest, I don't know. Here's the kind of thing that I was referencing though:
https://youtu.be/tj1dFIUAs7A?si=a4KYg6QEaxALHnWX&t=207
I've always thought Bill Gates was pretty smart, smarter than me, and has dedicated a lot of his time to researching things like pandemics. More time than I have dedicated to such research.
Seems like Meta's release of Muse Spark Open does not move the needle on things like this, since GLM-5.3 was already released back in August.