6 ms·
Idk if this work really bears on that argument. I can imagine Bender reacting to this by observing that this is a clever way of finding the needle (a good heuri
by lgessler 3y ago
Idk if this work really bears on that argument. I can imagine Bender reacting to this by observing that this is a clever way of finding the needle (a good heuristic algorithm for an NP hard problem) in a haystack (millions of other heuristic algorithms that are bad, very bad, or maybe not even syntactically well-formed), and then saying that the fact that the model still produced so much garbage is proof that the LLM still doesn't really know how to reason or access meaning.
And I think that'd be a good argument. OTOH, this system is not just an LLM, but an LLM and a bunch of additional components on top of it, and there might be different things to say about this system considered as a whole.
- airstrike 3y agoWe don't need LLMs to reason or to access meaning for them to be useful
- lgessler 3y agoSure, but I don't think even Bender or Marcus would deny their potential for practical utility, right? I think they mean to say that LLMs are not exactly as miraculous as they might seem, and very capable of misbehavior.
- cbb330 3y agoyour mind also conjectures many unreasonable and illogical solutions to things. then separately your mind has an "evaluator" to deduce a set of solutions down to things that make sense. Is anyone saying LLMs in isolation and without contextual tooling are miraculous? Even the text box for chatGPT is a form of wrapping LLM in a UX that is incrementally more miraculous.
- janalsncm 3y agoBender and Marcus don’t get to decide what can be useful for people. LLMs are useful, although their imperfections limit their usefulness. But the bar for usefulness is a lot lower than “perfection”.
- empath-nirvana 3y agoI'm not sure why people always want to consider the LLM in isolation. Like if someone built an apparently fully sentient android which completely mimics the behavior of a sentient human in every respect, but it was comprised of a bunch of LLMs and other systems wired together, would you say that it's not intelligent because the LLMs aren't fully sentient?