CamperBob2

Wartime progress in everything from radio to radar to rocketry has led directly to modern communications tech including satellites and microwave links.

When WW2 started, a "computer" wore a dress to work and was "programmed" to watch meters, turn knobs, do a lot of arithmetic by hand, and avoid asking questions. When WW2 ended a "computer" was a machine with a rapidly-increasing number of uses.

Before WW2, aircraft used propellers, flew slow and low, and crashed a lot. After WW2, the jet age ramped up practically overnight thanks to experience gained by all sides.

And yes, we got pesticides, herbicides, synthetic fuel and rubber processes, mass-produced antibiotics, probably a bunch of other stuff I'm not thinking of.

Then there's the whole Pax Americana thing, which is the part we're busy throwing away right now. It led to a relatively peaceful, stable global environment in which new technologies could emerge and spread without the need for a world war to inspire them. Now that America has basically been subverted from within, we'll soon get to find out whether that was a good thing for the world or not.

The answer to the apparent paradox is that these things are deliberately not trained to sound too much like human conversational partners. The labs don't want the bad press that they'd get from people creating deceptively-convincing bots, or from people forming emotional bonds with them like they did with GPT-4o. If you actually RLHF'ed a frontier-grade LLM to pass a Turing test, rest assured, it could do it.

"It's not x, it's y" and other goofy superficial tells do not have to be part of an LLM's response. But the last thing OpenAI wants to release is a GPT-4o with twice the IQ, so we have to put up with a lot of stupid clanker clichés.

For evidence, just look back at the best conversational models from a couple of years ago, and you will probably agree that they are better at fooling humans than their newer counterparts are.

A society where 98% end up unemployed because they no longer need to work for a living is something else completely.

RTL-SDR is the cheapest imaginable commodity implementation of SDR technology, while USRP was... well, pricy enough to bring in some solid revenue and get his company bought. More power to him, it was (and is) an incredibly useful platform.

There is still a lot we don't know about how to get the most out of existing LLM components from a speed or cognitive-performance perspective. People could easily spend the next decade studying and refining what's been built so far, even if no new, original approaches ever arrive.

WW2 ultimately brought massive wins for everybody who didn't die in it or otherwise suffer directly. We are now on the verge of throwing away many of those wins, but that's neither here nor there.

I swear, the audience around here these days...

So why in the world wouldn't I respond to this by automating away the hassle?

The solution for both sides is for the companies to implement support bots that have more power to address my problem than the humans they replaced were ever given. I have a feeling it's not going to work out that way, though. Maybe with a few unusually-enlightened organizations, but I'm not holding my breath.

At the same time, when I need to use the hardware for something, whoever is renting it from me at the moment is going to get unceremoniously booted, and I imagine they are not going to be happy about that. I assume that vast.ai's providers get uptime ratings that drive their work allocation, right?

Meanwhile, in the US, companies like Boeing, GM, and Intel will never be allowed to experience more than minor financial inconvenience before the government bails them out with protectionism, guaranteed loans, and outright subsidies.

I don't see a material difference between how the Chinese government treats their strategically-important industries and the way we do here in the West. Terms like "socialism," "Communism," and "capitalism" are just fodder for Fox News camp-followers.

The AI models we currently have still maintain their working state with a finite and laughably-small context window, but that is already starting to change. My .claude directory contains over 300 .md files that I didn't put there myself. Their contents are eye-opening. The question of who owns, stores, maintains, and can access that data is going to become insanely important over the next couple of years.

If you thought LLMs themselves were disruptive and contentious, just wait until the fight over object permanence gets under way. That's when owning your own box full of graphics cards is going to become important. My own bet is that I won't care too much about the electric bill or my opportunity cost when we all find out what the AI labs really have in mind, and what they're going to have to do in order to justify the valuations they're seeking.

TL,DR: it's not about the tokens, IMHO.

It's worth it, knowing that there are no rugs Sam or Dario or anyone else can pull.

So if you can spend $30K and immediately start mining $1500/month out of thin air, that's a pretty nice investment even if the electricity costs $200/month. Two years later the cards will have paid for themselves entirely and (I suspect) will still be pretty useful.

It does argue in favor of just paying OpenAI or Anthropic for tokens, though.

Not $400 a month, it doesn't. At least not around here (we average $0.12/kWh and I don't personally run my cards over 300W.)

And there's a lot of labour and effort involved...

Theoretically, the people who hang around HN are more likely than most to be capable of the labour and effort of setting these things up and maintaining them.

People don't even run their own email servers...

People don't run their own email servers because a convenient coalition of spammers, standards bodies, and large email providers have done their best to make running one's own email server almost impossible.

Interesting, you don't find that the model is perfectly capable of solving those on its own?

Or does it just refuse to try?

How do you think Trump has made more profit during his years in the Oval Office than he did in 30+ years of business before that?

And don't bother trying to claim that they cheated. Even if they did crib off of other mathematicians' notes, (a) that's how this is done; (b) those mathematicians were relying on AI as well; and (c) OpenAI resolved an aspect of the N-S problem that was not being addressed by anyone else.

One-shotting something like Pac-Man doesn't prove much. At the end of the day, one-shot fidelity is going to scale more or less linearly with model size/world knowledge. Why wouldn't it?