[gfodor]

I’m saying that seeing dubious economics is blinding people from accepting what’s actually going on with neural networks, and it leads to them having a profoundly miscalibrated mental model. This is not like analyzing a typical tech cycle. We are dealing with something here that we don’t really understand and transcends basic models like “it’s just a really good tool.”

[saubeidl]

That pseudo-religious angle that seems to have infected a lot of the tech industry is part of what "doomers" like myself or Zitron criticise.

It is just a really good tool. And that's fine. Really good tools are awesome!

But they're not AGI - which is basically the tech-religious equivalent to the Second Coming of Christ and about as real.

The fear isn't about the practicability of the tool. It's about the mania caused by the religious component.

[tim333]

I've followed the human level intelligence stuff for about 45 years, back before it was called AGI and the basic thesis is kind of anti religious. It's that human intelligence is basically the result of a biologically constructed computing device, not some god given spirit, and as human built computing devices continue their Moore's law like progression they will overtake at some point.

It's been true and kind of inevitable since Turing et all started talking about it in the 1950s and Crick and Watson discovered the DNA basis of life. It's not religious, not a mania, not far fetched.

[saubeidl]

The angle currently is the opposite. They're positing that the machine has some sort of spirit - see the other poster talking about "unexplained emergent intelligence".

[gfodor]

Saying we don’t understand why LLMs are intelligent is both true and completely unrelated to religion. You inserted the word “spirit” so perhaps you are the one conflating the two.

[tim333]

Well it's sort of true in that people stick these LLMs together and they produce intelligent seeming outputs in ways that the people building them don't fully understand. Kind of like how evolution stuck a bunch of biological neurons together without needing to fully understand how it works.

[gfodor]

It’s not a religious angle, we literally don’t know how or why these models work.

Yes we know how to grow them, but we don’t know what is actually going on inside of them. This is why Anthropic’s CEO wrote the post he did about the need for massive investment in interpretability.

It should rattle you that deep learning has these emergent capabilities. I don’t see any reason to think we will see another winter.

[jstanley]

100% agreed, but that is not a reason to spend this much money on it.

[saubeidl]

We don't know how acetaminophen works exactly either, but it's still just a really good tool...

[gfodor]

If Acetaminophen started exhibiting signs of human level intelligence I would have a similar perspective

[B56b]

Pretty obvious that your opinions have "nothing to do with the financials".

[gfodor]

Correct, my opinions have nothing to do with financials. If I was around for the discovery of fire I wouldn’t be wondering about the impact on the bottom line.

[jstanley]

Ah, this time it's different. Understood.

(To be clear, I do agree that AI is going to drastically change the world, but I don't agree that that means the economics of it magically make sense. The internet drastically changed the world but we still had a dotcom bubble.)

[gfodor]

Yep, this time it’s different.