ordersofmag

I read it in a cabin on a lake, far from any internet access, and without a computer. I still found it quite useful.

I think it's more accurate to say that the fundamental differences (between LLM's and humans) are abundantly clear and what we haven't really got a handle on are the similarities. Why is it that an LLM's can so convincingly mimic some things that humans do? Of course the fact that we don't really understand how humans do it in the first place is a real barrier. Perhaps looking at the intersection will shed some light on how humans work.

There is no such thing as a GPS radio in a phone in the sense of something that broadcasts a signal. GPS signals are broadcast by satellites and any device on earth can choose to listen for the signal and triangulate their position by receiving the signal from multiple satellites. GPS doesn't send out any signal from your phone to work. In the same way that your eyeballs don't need to emit light for you to see (unless you're a superhero). Phones broadcast signal to connect to cell networks and wifi networks.

Now if you expect your phone to show your position on a map (using the GPS coordinates it calculated from the satellite signal) then it likely needs to retrieve map data from the internet (and in doing so it might be broadcasting over cell or wifi frequencies). But that's a different story.

Much confusion is caused by people saying "I tracked X with GPS" with the implication that GPS allows tracking of remote objects. It doesn't. It allows the remote object to know its own location. For that information to get to someone else (the spy in the chasing vehicle) the remote object needs to be broadcasting that location information through some other means (which has nothing to do with GPS).

True for the pre-training data scraped from diverse sources. Less so for the later stage data for RL which is by all account more of a differentiator. In most cases the labs themselves produced the data so they are the copyright holders (or they are borrowing it from other labs via distillation). A lot of it is synthetic data, and since you can't copyright AI output, it becomes less about copyright and more about trade secrets.

Historically large data centers used evaporative cooling. So the water disappeared in sense that it's no longer available in a useful form in it's original locations. That's changing with growing use of closed-loop systems.

True if you're using a closed-loop system. But if you're using evaporative cooling (which is the main cause for concern) then the place where the water 'comes back into the environment' may be quite far away (at least in a useful sense).

This is entirely dependent on where you are geographically in terms of rainfall and weather. Both water and energy are abundant. The trick is getting them where you need them. So the where is what matters.

Please compare 1 million gallons to the 1 quadrillion gallons in the aquifer, or the 31 trillion that falls from sky each year in Nebraska and get back to me.

I think you confused "can't" with "shouldn't".

How to say you've never met a physicist....

These are all silly non-problems (speaking as someone who works with lots of professors around how to put AI systems to effective use)

-the thing that motivates you doesn't have to be the same thing that teaches you (e.g. a book can be a useful learning tool even if it isn't human)

-you can absolutely choose to interact with an LLM in ways that generate productive struggle. If you naively ask it for answers it won't be much use in this regard. But that's just a poor choice of how to use the tool. Just like hitting yourself on the head with a book doesn't mean the problem is with the book as learning tool.

-You don't need structured content 'on long scales' to learn. You learn a little bit at a time. Long textbooks are a side effect of the economics of moving paper around. Not a prerequisite for learning. An LLM will happily mimic the table of contents of a textbook or an outline for course. And you can use future prompts to generate the bits you need as you need them. One-shoting a textbook is a silly use of the tool. Plus, we're not short on textbooks. We're short on ways students can get customized help: answers to the specific ill-posed questions they have in their heads. LLM's are quite good at this. You can have textbooks AND LLM's (and professors)

-LLM's are happy to disagree with you. Just ask them to and they oblige. Ask them to respond as if they were a tutor and they will mimic the assuredness if appropriate.

Spoken as someone who has never seen a weather forecast. Computers have long produced results that were, if interpreted naively, incorrect. These systems require humans to take into account the conditional, probabilistic nature of the output and the inherent limitations of the model the computer was executing, when they interpret the output. LLM's are no different. They give you next token prediction probabilities. That is all. There is no particular reason to think that comes (or even could come) with a promise that those tokens will express truths about the world and only truths. The fact that greed-driven hype, and our human inclination to anthropomorphize the token generator inclines us to imagine things should be different doesn't make it so. No model output is truth. Never has been.

I'm genuinely unclear on what you mean here. Do you mean: 1. Most mathematicians are currently largely paid as part of their academic job (as a faculty member) and will continue to do that work (including the research part) even if extra-mural grant funding goes away. That seems entirely plausible to me.

Or do you mean? 2. Real mathematicians will continue to devote a significant part of their life/energy to mathematics research as a hobby, even if no one is paying them (in any way) to do it? That seems more of a reach and makes me wonder if many currently folks current employed as mathematicians aren't 'Real mathematicians' from your point of view.