Ask HN: Are people in tech inside an AI echo chamber?

It will take longer than some "evangelists" anticipate, but other than a lot of tech hypecycles (like web3), this one is different for 3 reasons:

    a) It's firmly in the public mindspace, including mass media
    b) It has already proven it's usefulness and applicability to real world problems
    c) Politics are taking it very seriously

A large part of that drive is due to the simple accessibility: Making ChatGPT as a convenient and simple webapp that appeals to non-techies was a brilliant move, and Microsoft driving integration into it's product suite will further adoption as well.

There are plenty of folks who don't fall into this trap.

Example: Me. I am fully aware that LLM generated code can be wrong or worse, contain disastrous errors. But being aware of that, and looking for the things I know from experience it commonly gets wrong, allows me to use it as an incredibly powerful tool in my workflow.

"Hallucination" is a term that works well for actual intelligence - when you "know" something that isn't true, and has no path of reasoning, you might have hallucinated the base "knowledge".

But that doesn't really work for LLMs, because there's no knowledge at all. All they're doing is picking the next most likely token based on the probabilities. If you interrogate something that the training data covers thoroughly, you'll get something that is "correct", and that's to be expected because there's a lot of probabilities pointing to the "next token" being the right one... but as you get to the edge of the training data, the "next token" is less likely to be correct.

As a thought experiment, imagine that you're given a book with every possible or likely sequence of coloured circles, triangles, and squares. None of them have meaning to you, they're just colours and shapes that are in random seeming sequences, but there's a frequency to them. "Red circle, blue square, gren triangle" is a much more common sequence than "red circle, blue square, black triangle", so if someone hands you a piece of paper with "red circle, blue square", you can reasonably guess that what they want back is a green triangle.

Expand the model a bit more, and you notice that "rc bs gt" is pretty common, but if there's a yellow square a few symbols before with anything in between, then the triangle is usually black. Thus the response to the sequence "red circle, blue square" is usually "green triangle", but "black circle, yellow square, grey circle, red circle, blue square" is modified by the yellow square, and the response is "black triangle"... but you still don't know what any of these things _mean_.

When you get to a sequence that isn't covered directly by the training data, you just follow the process with the information that you _do_ have. You get "red triangle, blue square" and while you've not encountered that sequence before, "green" _usually_ comes after "red, blue", and "circle" is _usually_ grouped with "triangle, square", so a reasonable response is "green circle"... but we don't know, we're just guessing based on what we've seen.

That's the thing... the process is exactly the same whether the sequence has been seen before or not. You're not _hallucinating_ the green circle, you're just picking based on probabilities. LLMs are doing effectively this, but at massive scale with an unthinkably large dataset as training data. Because there's so much data of _humans talking to other humans_, ChatGPT has a lot of probabilities that make human-sounding responses...

It's not an easy concept to get across, but there's a fundamental difference between "knowing a thing and being able to discuss it" and "picking the next token based on the probabilities gleaned from inspecting terabytes of text, without understanding what any single token means"

No it isn't, and I say that as a user and Integrator of ML tech.

Climate Change: Literally our planetary habitat becoming uninhabitable for our species.

Russia: A country with one of the largest nuclear arsenals literally threatening a large portion of the world with nuclear annihilation.

Excuse me when the prospect of changes in the economic landscape for white collar jobs doesn't look particularly frightening compared to such problems. Especially since we live in a day and age where 2 entire generations have lived most of their lives in almost constant economic upheavals anyway.

As for all the AI-doomerism that's flying around the net: As long as no one can even give me a precise, quantifieable definition of "general intelligence", aka. one that doesn't include pointing at ourselves, and a method to measure how far AI is from that, I will work under the assumptions confirmed by what is measureable and observable: that what we have are still stochastic inference engines.

The DLR for example still has drivers, its just they are located somewhere other than the train.