A fairly lengthy article about Autonomous AI, and, as far as I can tell, not a single word about the safety implications of such a system (a short note about reliability of LLMs is all we get, and it's not clear that the author means anything more than "the thing might break").
I get that there are different philosophies on AI risk, but this is like reading an in-depth discussion about potential atomic bombs in 1942, with no mention of the fact that such a bomb could potentially level cities and kill millions.
If a field of research ever needed oversight and regulation, this is it. I'm not convinced that would solve the problem, but allowing this kind of rushing forward to continue is madness.
I get that there are different philosophies on AI risk, but this is like reading an in-depth discussion about potential atomic bombs in 1942, with no mention of the fact that such a bomb could potentially level cities and kill millions.
If a field of research ever needed oversight and regulation, this is it. I'm not convinced that would solve the problem, but allowing this kind of rushing forward to continue is madness.