[amelius]

Using an LLM doesn't mean it has to take the final decision. You can also use it as a warning system.

[Mawr]

False negatives are a huge issue when designing safety systems. It is not the case that "more warnings = more better".

[amelius]

Of course, but an LLM can potentially help with that.

[stnikolauswagne]

Is there any indication that current warning systems are insufficient in any way that would be improved by LLM involvement?

[captainbland]

Well they don't attract nearly as much investment in the current market, I think that might be the problem people really want to solve

[vidarh]

We won't know that until someone has actually investigated how an LLM would do in those scenarios.

[stnikolauswagne]

That sounds like a solution looking for a problem though, i see plenty of arguments against throwing critical safety information that are in charge of peoples lives into an LLM "just in case the result is better than the result that the current battle-hardened systems already provide"

[amelius]

Nobody can be against just collecting the data and letting people experiment with it.

[stnikolauswagne]

Are all those security systems actually open right now? Because that sounds like an absolute security nightmare if so.

[amelius]

Can you give an example scenario?