3 ms·

As soon as we started unironically calling LLMs "AI" we went down the hype path. That has plenty of downsides, like stressing out the entire world and attractin

As soon as we started unironically calling LLMs "AI" we went down the hype path. That has plenty of downsides, like stressing out the entire world and attracting cryptocurrency bros, but also the major upside massive of funding/acceleration. So far, all we have is more software running on computers. It's powerful, and it's amazing, but it's not magic. Calling it "AI" was possibly a net-negative but we don't know yet.

cautiouscat 3mo ago
> Calling it "AI" was possibly a net-negative but we don't know yet. I’m not sure it’s net negative or not. I’ve found that it’s reductive though. We have this really broad field of artificial intelligence reduced down to at worst a “slop machine” and at best a single tool. Imagine being a CS professor that studied AI in the 90s and how you have to over explain you don’t mean LLM chatbots to a layman.
lukan 3mo ago
"It's powerful, and it's amazing, but it's not magic" But since its creators and as of my knowledge everyone else totally did not see it coming, that you can now give a vague prompt full of spelling errors - and get returned a working program - I would say it is pretty close to magic (as in we don't really understand why it works so good). I also don't see how you cannot call it AI. Especially since simple chess engines and alike were called AI long ago. So it is not general strong AI and has no consciousness and no mind and is pretty dumb too often - but the general concept - getting from a some vague text to a working program has some connection to intelligence to me.
- jacobgold 3mo ago
  Yes, LLM agents are "magic" in the sense that "any sufficiently advanced technology is indistinguishable from magic"[1] But it's not actually magic. Technical people understand that it's just software running on computers. 1. https://en.wikipedia.org/wiki/Clarke%27s_three_laws https://en.wikipedia.org/wiki/Clarke%27s_three_laws
  - lukan 3mo ago
    Sure, nothing is magic. You can go look how a simple LLM works and build your understanding from there. But calling it "just software" is trivializing it in my opinion. I can write software, but I cannot write software that writes software.
    - jacobgold 3mo ago
      > But calling it "just software" is trivializing it in my opinion The bigger mistake would be trivializing the rest of the technology involved just because LLMs are the newest piece. LLMs are only "magic" because they're built on a stack that was already "magic" without them. LLMs are impossible without: - operating systems - programming languages - compilers - data centers / power grids / air conditioning - servers / switches / routers - CPUs / RAM / GPUs / SSDs - fiber networks - etc
      - int_19h 3mo ago
        I would argue that LLMs are still magic, because, unlike the rest of this list, we still don't actually know how exactly they do what they do. We know how to build and train them, and we have general ideas about why e.g. attention is important, but if we actually knew how they tick we could do the same thing in a much smaller package (by handcoding everything).
nylonstrung 3mo ago
I think calling it AI has been very negative. One of the lesser, but still underdiscussed ramifications is that I think it has limited the public's ability to comprehend the Yann LeCunn argument, that genuine AI is likely possible but that LLMs and transformers are a dead end and we need to explore different modalities

cautiouscat 3mo ago

> Calling it "AI" was possibly a net-negative but we don't know yet. I’m not sure it’s net negative or not. I’ve found that it’s reductive though. We have this really broad field of artificial intelligence reduced down to at worst a “slop machine” and at best a single tool. Imagine being a CS professor that studied AI in the 90s and how you have to over explain you don’t mean LLM chatbots to a layman.

lukan 3mo ago

"It's powerful, and it's amazing, but it's not magic" But since its creators and as of my knowledge everyone else totally did not see it coming, that you can now give a vague prompt full of spelling errors - and get returned a working program - I would say it is pretty close to magic (as in we don't really understand why it works so good). I also don't see how you cannot call it AI. Especially since simple chess engines and alike were called AI long ago. So it is not general strong AI and has no consciousness and no mind and is pretty dumb too often - but the general concept - getting from a some vague text to a working program has some connection to intelligence to me.

jacobgold 3mo ago
Yes, LLM agents are "magic" in the sense that "any sufficiently advanced technology is indistinguishable from magic"[1] But it's not actually magic. Technical people understand that it's just software running on computers. 1. https://en.wikipedia.org/wiki/Clarke%27s_three_laws https://en.wikipedia.org/wiki/Clarke%27s_three_laws
- lukan 3mo ago
  Sure, nothing is magic. You can go look how a simple LLM works and build your understanding from there. But calling it "just software" is trivializing it in my opinion. I can write software, but I cannot write software that writes software.
  - jacobgold 3mo ago
    > But calling it "just software" is trivializing it in my opinion The bigger mistake would be trivializing the rest of the technology involved just because LLMs are the newest piece. LLMs are only "magic" because they're built on a stack that was already "magic" without them. LLMs are impossible without: - operating systems - programming languages - compilers - data centers / power grids / air conditioning - servers / switches / routers - CPUs / RAM / GPUs / SSDs - fiber networks - etc
    - int_19h 3mo ago
      I would argue that LLMs are still magic, because, unlike the rest of this list, we still don't actually know how exactly they do what they do. We know how to build and train them, and we have general ideas about why e.g. attention is important, but if we actually knew how they tick we could do the same thing in a much smaller package (by handcoding everything).

nylonstrung 3mo ago

I think calling it AI has been very negative. One of the lesser, but still underdiscussed ramifications is that I think it has limited the public's ability to comprehend the Yann LeCunn argument, that genuine AI is likely possible but that LLMs and transformers are a dead end and we need to explore different modalities