alt Hacker NewsI don't think so - they could get a lot better, but it's going to take a lot of human brainpower to get there. They are barely adequate now, but their weaknesses are sold as intrinsic to the nature of the software in a way that mere civilians are too dumb to understand.
But imagine a world in which you could ask AI to write software to accomplish a particular task, and it understood completely what you meant, made tasteful choices, interviewed you as it went along to clarify things, and generated beautifully commented (maybe even Knuth-style literate) code that wasn't dogshit. Doesn't seem like an impossibility. What is annoying is that we're pretending like we have anything like that now.
Instead, there's this sort of science fiction pitch where we pretend like the LLM is just too damn logical to understand your sloppy human language. It's not logical, it's built on (human-generated) slop and generates slop. It's simply failing. We've come up with better ways to orient it towards big picture thinking and filling in outlines, but we don't know how to train it on big pictures.
Not that this would stop open models/harnesses from following (or even leading) frontier labs and destroying their insane valuations, but I don't think there's any evidence that a plateau has been reached. What we have is not great, and while a wall may have been hit on current approaches, the money, compute, and simply current progress in ideas has us in an environment where everything could change over the course of a crazy day. A single blog entry.
I do think that the frontier labs seem to have reached at least a temporary plateau. But if they figured out how to move past it and this new method didn't require a lot more compute, they'd still be screwed. Their only out is to pass the bag to the public, but Firmus, CharacterAI, CoreWeave...
Replies
> generated beautifully commented (maybe even Knuth-style literate) code that wasn't dogshit. Doesn't seem like an impossibility.
It seems closer to impossible than possible with the technology we have. Maybe something that isn't an LLM will come around that can do it, but even if that something else was capable of delivering perfect, ordered, maintainable, code without bugs and hallucinations it would have to be practically magical to understand completely what people mean when they talk about what they want because people are incapable of articulating what they want in a way that even other humans can understand.
For anything non-trivial, a technology capable of getting that information out of person without countless meetings would have managed to accomplish something far beyond what entire teams of humans are able to do today.
> It's not logical, it's built on (human-generated) slop and generates slop.
This is why if we haven't hit a plateau yet we very quickly will. All the good human generated slop has already been consumed and is being drowned out by vast amounts of the lossy, error filled, and highly concentrated slop AI produces from the human slop it was trained on.
Do you really think an LLM that spews out content based on statistical relationships is going to improve when it's been primarily trained on the slop of previous generations of LLMs? Hallucinations and errors will carry forward and spread like cancer. All the intelligence and creativity LLMs borrowed from human output will homogenize as the LLMs distill and then reiterate, effectively making bullet points out of bullet points, until their mimicry of intelligence becomes unintelligible.
Y'all are overthinking this: having successfully laundered IP, then successfully sold IP-laundering-as-a-service, then successfully laundered liability, they now want to sell liability-laundering-as-a-service.