Re:What Is AGI-Hard
by mannykannot on Thursday January 05, 2023 @09:15AM
(#34259807)
What you call a huge opportunity is actually (or also, if you prefer) a major shortfall in the current LLMs. Whether it will turn out to be merely temporary, and more pertinently, how it might be solved and whether it can be solved with more of the same, are themselves unanswered questions at this time.
Right, it’s a fundamental disconnect between how an autoregressive statistical language model generates text and how people think about talking and writing. We have known this for a long time. It is the essence of the civil war between symbolic NLP folks and statistical language modeling folks. Now that ChatGPT is quasi-mainstream, the strengths and weaknesses of the latter are on display for all to see.
My only point is that if we’re approaching a local maxima with purely statistical approaches, it raises the stakes for symbolic approaches to come through.
> This is a big problem now that there are systems able to produce huge volumes of plausible blithering that's factually wrong.
If because of some law of nature ChatGPT were the last iteration of automatic text generation we could ever create, it would be a big and permanent problem. Right now it’s just a big and temporary problem, with a huge opportunity for automatic fact-checking. (Or more precisely, validity and soundness checking.) Maybe the answer to the bullshit factory is a rebalancing of Brandolini’s law.