Hacker News Reader

Comment by jcgrillo | Hacker News Reader

This is absolutely on point, I don't understand why your post is being down-voted.

1. It's not actually refuting the point being made which is "it's seems hard to take advantage of LLMs' capabilities", which to me seems like a good point that stands on it's own, regardless of who is saying it.

2. The original post is, to my eyes, not implying anything about training a model to help use other models, so the second part seems irrelevant.

I disagree. The comment is striking directly at the claim that people aren't taking "advantage of LLMs' capabilities". This is their capability and no amount of "clear communication" is going to change that

One of the rules of HN is to assume good intent. People do and say the right thing often even when it's antithetical to their source of income. If there is a substantive way to otherwise then say it, don't immediately write-off people's opinions, especially people in their fields because of a possibility of bias.

OK, but the post this was in response to was bloviating about all kinds of sci-fi stuff like "super-intelligence" and the like. It was the opposite of "antithetical to their source of income", instead it was playing into some techno-futurist faith cult.

There is a strong argument that super-intelligence is already in the rear-view mirror. My computer is better than me at almost everything at this point; creativity, communication, scientific knowledge, numerical processing, etc. There is a tiny sliver of things that I've spent a life working on where I can consistently outperform a CPU, but it is not at all clear how that could be defensible given the strides AI has made over the last few decades. The typical AI seems more capable than the typical human to me. If that isn't super-intelligence then whatever super-intelligence is can't be far away.

> given the strides AI has made over the last few decades

This is where I lost the plot. The techno-futurists always seem to try to co-opt Moore's law or similar scaling laws and claim it'll somehow take care of whatever scifi bugaboo du jour they're peddling, without acknowledging that Moore's law is specifically about transistor density and has nothing to do with "strides" in "AI".

> whatever super-intelligence is can't be far away.

How do you figure? Or is it just an article of your faith?

It's been the same old, tired argument for seven decades--"These darned computers can count real real fast so therefore any day now they'll be able to think for themselves!" But so far nobody's shown it to be true, and all the people claiming it's right around the corner have been wrong. What makes this time different?

> How do you figure?

https://epochai.org/blog/training-compute-of-frontier-ai-mod...

https://ourworldindata.org/grapher/test-scores-ai-capabiliti...

We're still seeing exponential upswing in compute and we appear to already be probing around human capacity in the models. Past experience suggests that once AIs are within spitting distance of human ability they will exceed what a human mind can do in short order.

These are not subtle trends.

I'm not sure what the amount of computer time spent on training the models has to do with anything, the article states it "is the best predictor of broad AI capabilities we have" without attempting to defend the claim. The "humies" (to use a dated term) benchmarks are interesting but clearly not super indicative of real world performance--one merely has to interact with one of these LLMs to find their (often severe) limitations, and it's not clear at all that more computer time spent on training will actually make them better.

EDIT: re: the computer time metric, by the same token shouldn't block chains have changed the world by now if computer time is the predictor of success? It makes sense for the industry proponents of LLMs to focus on this metric, because ultimately that's what they sell. Microsoft, NVidia, Google, Amazon, etc all benefit astronomically from computationally intensive fads, be it chatbot parlor tricks or NFTs. And the industry at large does as well--a rising tide lifts all boats. It's not at all obvious any of this is worth something directly, though.

> I'm not sure what the amount of computer time spent on training the models has to do with anything

Fair enough. What do you think is driving the uptick of AI performance and why don't you think it will be correlated with the amount of compute invested?

The limitations business looks like a red herring. Being flawed and limited doesn't even disqualify an AI from being super-intelligent (whatever that might mean). Humans are remarkably flawed and limited, it takes a lot of setup to get them to a point where they can behave intelligently.

> EDIT: re: the computer time metric, by the same token shouldn't block chains have changed the world by now if computer time is the predictor of success?

That seems like it would be a defensible claim if you wanted to make it. One of the trends I keep an eye on is that log(price) is similar to the trend in log(hash rate) for Bitcoin. I don't think it is relevant though because Bitcoin isn't an AI system.

You forget energy efficiency friend. We are FAR more efficient...

> The typical AI seems more capable than the typical human to me

Your microwave is more capable than a typical human.

If of course your definition of capabilities is narrowly defined to computations and ignores the huge array of things humans can do that computers are no where close to.

> the huge array of things humans can do that computers are no where close to.

For example?

anything involving an interface with the physical world.

For example, running a lemonade stand.

You'd need thousands, if not millions of dollars to build a robot machine with a computerized brain capable of doing what a 6 year old child can do - produce lemonade from simple ingredients (lemon, sugar, water) and sell it to consumers.

Same with basically all cooking/food-service and hospitality tasks, an physical therapy type tasks (massage, chiropractor, etc)...

heck, even driving on public roads still doesn't seem to be perfect, despite 10+ years on investment and research by leading tech companies, although there is also a regulatory hurdle here.

You seem to have shifted the conversation's goalposts there - those are things that computers can do, it just costs a lot.

And, more to the point, they aren't indicative of intelligence. Computers have cleared the intelligence requirements to run a lemonade stand by a large margin - and the other tasks too for that matter.

> those are things that computers can do, it just costs a lot

One could travel between continents in minutes on an ICBM with a reentry vehicle bolted to the front but we don't because it's too expensive. It's a perfectly reasonable constraint to demand that a technology be cost effective. Otherwise it has no practical value.

>My computer is better than me at almost everything at this point; creativity, communication, scientific knowledge

By that logic, libraries, art museums and calculators are "super intelligent"

Are you in the habit of talking to your calculator?

Reply on news.ycombinator.com