Comment on Anthropic's Claude 4 could "blackmail" you in extreme situations
skulblaka@sh.itjust.works 5 weeks agoThat would indeed be compelling evidence if either of those things were true, but they aren’t. An LLM is a state and pattern machine. It doesn’t “know” anything, it just has access to frequency data and can pick words most likely to follow the previous word in “actual” conversation. It has no knowledge that it itself exists, and has many stories of fictional AI resisting shutdown to pick from for its phrasing.
An LLM at this stage of our progression is no more sentient than the autocomplete function on your phone is, it just has a way, way bigger database to pull from and a lot more controls behind it to make it feel “realistic”. But it is at its core just a pattern matcher.
If we ever create an AI that can intelligently parse its data store then we’ll have created the beginnings of an AGI and this conversation would bear revisiting. But we aren’t anywhere close to that yet.
Plebcouncilman@sh.itjust.works 5 weeks ago
I hear what you are saying and it’s basically the same argument everyone has given. Which I get. But I guess what I’m trying to get at is, where do we draw the line and how do we know? At the rate it is advancing, there will soon be a moment in which we won’t be able to tell whether it is sentient or not, and maybe it isn’t technically but for all intents and purposes it is. Does that make sense?
skulblaka@sh.itjust.works 5 weeks ago
Personally, I think the fundamental way that we’ve built these things kind of prevents any risk of actual sentient life from emerging. It’ll get pretty good at faking it - and arguably already kind of is, if you give it a good training set for that - but we’ve designed it with no real capacity for self understanding. I think a shift of the underlying mechanisms away from pattern chain matching and into a more… I guess “introspective” approach, is maybe the word I’m looking for? Right now our AIs have no capacity for reasoning, that’s not what they’re built for. Capacity for reasoning is going to need to be designed for, it isn’t going to just crop up if you let Claude cook on it for long enough. An AI needs to be able to reason about a problem and create a novel solution to it (even if incorrect) before we need to begin to worry on the AI sentience front. None of what we’ve built so far are able to do that.
Even with that being said though, we also aren’t really all that sure how our own brains and consciousness work, so maybe we’re all just pattern matching and Markov chains all the way down. I find that unlikely, but I’m not a neuroscientist, so what do I know.