bigbadfeline

There's a lot of free stuff incurring "real costs" in its production, one more or less doesn't make a shred of difference.

> I agree with you that producers of these models should be able to extract value from their models.

Nobody is stopping them from extracting value but they fail despite being placed in a very advantageous position vis-a-vis others, which tells me, they should call it quits or go open like everybody sane.

Open is the way to go, nothing else works in the field of AI, that's the bitter lesson closed labs are yet to learn.

Tell me you know nothing about economics without telling me you know nothing about it.

I do.

> Pol pot wasn’t a genius. But he also wasn’t a monkey.

Is that supposed to be a revelation?

> How many times in history has an animal successfully ruled over a population of humans?

So you consider AI to be qualitatively superior to human intelligence in the same way humans are superior to animals? It's a wild claim with no basis in reality and it tells me you know very little about both AI and humans.

Your perception of yourself as an animal next to what you consider "superior AI" isn't a joke though, I don't know how you got that depressed, but I hope there aren't many people like you, because that kind of self-denigration, fear and submissiveness can become a serious political problem.

Oh, the drama. Mr Parnell seems oblivious to the existence of system prompts, harness prompting and training data - any of these, alone or in some combination, could trigger "research" and together with lousy user prompting can do anything at all.

The problem is closed LLM providers, we don't know what they do on their end and inside their harnesses on ours.

Is this what you're talking about:

"Oracle has issued a force majeure notice regarding its New Mexico AI data center project, known as Project Jupiter. This legal move is intended to allow Oracle to delay payments if the project does not meet its expected completion date in 2028."

This has nothing to do with memory prices and everything to do with too much cash chasing too few physical goods used in construction and power generation, which is a systemic problem upstream of RAM overpricing.

> And you're trying to say someone should go out and invest 5 years

Did I say that? Where? I talked about Chinese silicon makers who work in a very different system. As to why we have a monopolized market, check [1].

> You arn't paying attention.

Attention is all you need - not me though, I've got enough of it. Check again the last paragraph of my previous comment to understand why.

[1] https://news.ycombinator.com/item?id=49927315

Exactly this. Unfortunately, this strategy has become standard practice in every area of the economy and it's a huge corrupting force.

There's nothing volatile about the AI bubble, it's been going on for a while regardless of cost and that will continue because of politics - which is rather transparent.

> I just don't see why someone who invest several multi billion dollars to squeeze into a potential window that may or may not last 3 years.

Interesting, Chinese companies don't seem to know this, they're investing bigly in RAM and other silicon despite being sanctioned and tariffed up the wazoo. Should we tell them that it's a crunch for no more that 3 years so they don't waste themselves for nothing?

The more interesting question is why Chinese RAM is hugely tariffed, if not for preserving the crunch and its monopoly pricing?

> I mean, we can assume the people who make the memory love getting huge prices for it.

Your mistake is to think of the RAM makers instead of the systemic reasons for the problem, these would have to include tariffs, sanctions, subsidies, etc that drive market pricing, RAM manufacturers and their potential competitors.

In economics, "demand" doesn't mean "I want this", it means paid, funded demand. 100s of millions of "free" LLM users aren't funded demand, they're subsidized. This leads to front-running and scalping the silicon market at an unreal scale, it also monopolizes access to silicon and compute. Nothing happens by chance here.

I don't think so, "in a computer" doesn't mean "in your computer". In many ways a dumpster is safer than a computer belonging to someone who likes to exclude you from access to valuable information - like the latest models of the AI bros, which are fully or partially blocked from the public but still available to selected few.

Wrong about what? That machines can do a lot of a harm if they are attached to critical infrastructure without supervision? That's always been the case and it's overcome by security provisions which can tame any machine, smart or not.

The people you enumerated aren't security specialists and they've never provided anything like a proper analysis of the security measures used by the perpetrators of hacking incidents - from prompts to harnesses to hardening to air-gaping, nothing is addressed. Moreover, they have glaring conflicts of interest, they belong to the already powerful who'd do anything to gain more power at our expense, they speak their stock portfolios, far from the whole truth.

I'm pretty sure a few script kiddies, with their own tools and without any LLMs, could have achieved the same breaches given the same freedom from consequences in the same environment. That latter point is the crux of the problem, the perpetrators of the AI breaches acted like script kiddies with carte blanche to cause harm and blame it on their tools, therefore whatever measures stop script kiddies will stop the AI bros too, there's no need for more power.

> It's frightening to be personally concerned about the possibility of this powerful technology getting out of control and to see little but conspiracizing in response.

A sequential word generator can be secured, nobody has provided any proof to the contrary, all we hear is empty generalities like "frighteningly powerful", we cannot trust the word of the perpetrators, that's not a sound security practice.

A distinction without a difference, the second automatically leads to the first, no exceptions.

> "everything for the government, above citizens". That is, forget the individual, there is a greater good for which we must endure some hardship, trust me.

Right, in addition to being a code for "we over the others", "nation above all" also provides the excuse for the social mobilization necessary for wars. Nationalism is a tool of war, it may start in some covert form but the end result is the same.

I don't think anyone argued differently, the actual question is, are humanities good for anything at their current state of development?

> You will be spending a lot of time working with other people, [on, and on, and on, how there's a world outside of technology]

That doesn't automatically qualify these humanities as some sort of aid. Looking at the social developments in the last few decades, things appear to be going downhill and accelerating. Humanities seem to be contributing to that trend, or at least completely unable to reverse or slow it down.

If people recognize something as being worthy of mockery and they are also aware of the situation, they gain nothing from reading some tripe that provides zero new information.

Also, a blog post shouldn't assume that all readers have seen the interview in question. What's the value of talking nonsense about something the readers have never been informed about? At most, the rational readers would be disgusted by the attempt to influence their opinion before they'd seen the facts.

> If the author needs to explain to you why not knowing your own address is foolish,

Well, the readers may not know that the mocked person doesn't know his own address, so when the author doesn't explain it, the author is the foolish one.

If the readers know the facts, mocking someone for trivialities without pointing out some novel reasons "why" is double foolish.

If you use AI, you lose expertise too! Retaining expertise isn't a one pony show, it can be done even when externalities work against it, shallow thinking can only do harm here.

Or maybe this is a veiled threat, now that they're messing around with viruses. This whole thing is getting worse by the hour.

A stochastic parrot with human knowledge, using human tools, running human-designed trial-and-error experiments can solve human-defined hard math problems. This is nothing new, automated mechanical proofs predate LLMs by many years, you being unaware of it doesn't change it.

That was before AI. You're acting like AI is some evil genius that can only attack and cause trouble but AI is a will-less tool directed by people, direct it to software safety and you will have software safety - cheap. Ditto for hardware.

> It's more like Russian roulette with 1 cylinder out of a million that's loaded.

Don't attach it to a gun then, ban offensive military AI, problem solved.

> The problem comes when you run it a few billion times a day, you'll shoot yourself in the face really quick.

Nonsense, run it a few "billion" times a day to fix vulnerabilities and you'll have no vulnerabilities. Do it in a safe environment... what's the big deal?

Theatrics aside, anticipating zero days isn't only possible, it's required, even for the unknown ones. It wasn't that long ago when the AI labs were spending millions of $$ running their models to find multiple vulnerabilities, they even argued that they don't have to follow responsible disclosure, so proud of themselves in their privileged hubris.

At that time, no hacking happened because the models didn't have access to the wide internet, they were confined to a local computer or cluster.

In the HF case their unaccountable hubris went even further - the engineers knew the models can find zero-days and escape, nevertheless they ran the "experiment" on a system attached to the internet - the hacking is entirely the fault of human engineers and managers.

Really?

> How many of our actions would look exactly the same.

Well, there are 8 billion of us, not exactly the same - a clear threat to humanity according to you.

> the statistical likelihood is all 21 of you would try to walk out to your car at 5 in one of the little loops that humans repeat every day.

This (statistically) doesn't happen with people because we follow rules. If you left your car on neutral and it struck another car in the parking lot, you broke the rules, not the car.

> LLMs are just an intelligence we can make a lot of copies of.

LLM's aren't intelligence, they are mechanical parrots of human knowledge. The meat parrots aren't all the same either and they don't repeat the same words for the same prompts, so what, ban parrots? If you left a bunch of parrots out of the cage, attached their beaks to gun triggers and they caused harm - you broke the rules, not the parrots.

Hey, it's teamwork, the teammates may be arguing and blaming each other but that's according to script too.

Don't bend over backwards to defend utterly oppressive regulations, simply increasing the de-minimums would reduce or eliminate package splitting, limiting the number of packs per shipping address would help too - there are many ways to solve this self-inflicted problem which comes from Byzantine regulations that use protectionism as a Trojan horse to achieve other sneaky objectives.

Billions and billions... of hearsay.

> and they have a unique and new tool that we as a species barely understand, and cannot defend against,

There's a causal link here, we cannot defend against something we don't understand, in this case more understanding and more open AI translate directly to better defenses.

> and no one seems to give a shit.

This is our problem and solving it involves a lot of politics, which, at this time, is an area where the good guys are in a seriously disadvantaged position, certainly a harsh test for those who engage in it.

I'm sure fresh numbers are available, they're just unpublished.

We could do some educated guessing here - in the last 5 years China's economy grew at the rates of 8.6%, 3%, 5.4%, 5%, 5%, the US grew at 6.5%, 2.5%, 2.9%, 2.8%, 2.2%. We can extrapolate the manufacturing numbers based on that, which would give some advantage to the US because the US manufacturing lags the average, while the Chinese leads it.

The math gives us the current values for

World: $19.4 T, (%6.5,%3.4,3x%2.9)

China: $6.3 T - 32.5%,

US: $2.9 T - 15%.