Neither of those look likely yet.

I'm pointing out that there's no known information theoretic constraint about the impossibility of frontier AI models being improved to fit/run on a small GPU.

Please do not make up plausible sounding science facts.

A Toyota dealership arranges a loan for you. Through a bank for a used vehicle, sometimes through Toyota itself for new cars.

A house builder will routinely take on part of the loan providing burden to get some of the interest.

Even someone selling you their thirty year old house will often provide seller financing.

You may have ideological opinions against this, which is fine. There are billions of people, for example that are fundamentally opposed to the idea of interest. But like it or not, this is how it has worked for the last ~500ish years.

It's almost a given that whatever is frontier intelligence today will run on a potato in a few years.

RTX 5070 prices go up ~N times. Nvidia makes more money because it's easier to make these things than it's to make a GB300.

You pay Apple for a MacBook, Apple uses it to develop a better MacBook.

What goes wrong is leverage. We haven't seen much hint of the 10x leverage kind of deals that brought down the house in 2008.

This will almost surely change with AI.

And what fraction of electromagnetic waves and particles are going to absorbed by those neutrons and other things in the air.

Pretty much weather!

It’s not just about out remembering, it’s about breadth.

Mathematicians are all about depth. It’s pretty much impossible to become an expert in more than one narrow field of mathematics.

AI is happily applying techniques and abstractions across these silos.

Parents of the kids, realistically.

It’s great to yearn for bug free software, but that necessarily brings an insane amount of red tape to get anything done. Like 5 years worth of review to start building your code with the next version of your compiler. Or six months to ship a hundred lines of code.

I'd rather have bugs. I’d happily quit having anything to do with software than suffer months of doing nothing in the name of security.

The signal here is the action of stopping the agent “do something else, this is stupid”, not a tweak to the initial prompt.

My comment was from a decade old perspective

Then the Google engineers who joined Facebook missed it so much that they built a better replacement.

As long as you are not missing important information, how you word the prompt does not have any effect.

It's mostly ad hoc scripts and some pretty horrible hacks being run by a hundred engineers trying to improve a thousand different things at once with a hundred thousand GPUs.

I'm sure once we understand the tech better, the training process will look like running a program.

Model training now is not a straight forwards process of input data -> run tools -> get model.

There's a whole lot of alchemy going on. We don't quite understand what works and what doesn't. Think of it like painting with water color and having to improvise very often.

The only advantage over water color is that we can revert to a working state.