Live data from Hacker News

The US is winning the AI race where it matters most: commercialization

avkcode.github.io

461–470 of 718 posts

Re: The US is winning the AI race where it matters most: commercialization

#461
post #277

No, they are not. They are winning because West is forbidden to use Chinese models for anything work-related.

> They are winning because West is forbidden to use Chinese models for anything work-related. Because the models hosted in China are not trusted. This is 100% a part of what makes up commercialization.

Is anyone outside the US trusting anything hosted in today's US? If so, why?

Re: The US is winning the AI race where it matters most: commercialization

#462
post #427

Earlier quoted context omitted.

But what do you do with all the waste energy? All those MW and GW have to end up somewhere and radiation into a vacuum is the hardest way to dump heat.

At 70-80C (working temp of silicon chips) 1m2 radiates 700-800W, i.e. the heat of 1 GPU like H200 without any need for any cooling equipment beside the radiator itself( and may be some dumb heatpiping) . To acquire that energy you'd need 3-4m2 of solar panels. So a datacenter would be a large field of solar panels with a smaller field of heat radiators in their shadow. To the commenter below: yes, exactly, this is wh…

You can probably (I haven't verified this) omit separate radiators and just use the back of the solar panels. Effectively you're describing mounting each H200 to the back of a 4 m^2 solar array at which point I suspect the equilibrium temperature will fall within an acceptable range. In fact the H200 and electricity are both entirely irrelevant here - the core question is what the equilibrium temperature of a black planar surface is at a given distance from the sun.

Re: The US is winning the AI race where it matters most: commercialization

#463
post #125

Earlier quoted context omitted.

Maybe I'm not doing capitalism right, but isn't it supposed to be "The one who profits the most wins"? If you win by just spending, I think you need to adjust the parameters of your capitalistic market.

I don't get it either, but it seems to me a bit similar to how the US, if you look at market value of car companies, has utterly crushed Europe and Japan (with China surging ahead of those and maybe threatening the US soon), which to me sounds crazy (I still think of German cars as the top of the bunch). According to Google (AI summary, no idea if it's 100% right but from what I've seen elsewhere it seems right): Top…

Look at profit rather than revenue - "it's easy to make a lot of revenue when you're selling a dollar for 80 cents" applies just as much to big legacy automakers as it does to startups.

Re: The US is winning the AI race where it matters most: commercialization

#464

Earlier quoted context omitted.

The USA did not take Iran. They essentially shot off fireworks, killed the figure head and then got check mated.

The U.S. isnt checkmated, the U.S. is enforcing a naval blockade against Iran and their oil based economy is in a free fall with rapid inflation. The U.S. is experiencing slightly higher gas prices but economy is still humming. Meanwhile the U.S. military has not exhausted all of its options while Iran has none.

The US is indeed checkmated like Afghanistan like Iraq and like Vietnam checkmated spinning your wheels spending money wasting money wasting time and wasting resources, checkmate.

With the addition of most countries now looking for other trade partners the Art of the no deal…

Re: The US is winning the AI race where it matters most: commercialization

#465

Considering that you can easily swap out one AI for another and there is zero lock-in potential, does it really matter who is winning now?

Government, industry, and general integration into the business software stacks are not easily swapped out. Most businesses are adding limitations on using open models.

> Government, industry, and general integration into the business software stacks are not easily swapped out.

My business's integration literally has a dropdown for which model you want to use. I think that's pretty standard.

Re: The US is winning the AI race where it matters most: commercialization

#466

Earlier quoted context omitted.

Yes, nonsensical people like EY don’t think it’s nonsensical.

Researches at top AI labs don't consider EY to be a kook even though they may not necessarily agree. EY concepts/terminology appear in Anthropic safety papers. Geoffrey Hinton takes him quite seriously and mentions him in his interviews.

Just because some researchers are infected with this idiocy that EY propagates does not mean that it is legit.

Maybe they should pay more attention to real problems like the sycophantic nature of current LLMs causing psychosis in people and worry less about theoretical AGI.

Re: The US is winning the AI race where it matters most: commercialization

#467
post #30

Anthropic, OpenAI and Google are the standouts, but the main question for me is, why is this a war? In their own context China has greatly benefitted from this. They shored up their gpu design and manufacturing expertise. If this really is a war, trump is kneecapping the country with his lawlessness and eroding America’s good will. If the world cannot trust China with their data and they cannot trust the U.S. to prov…

The idea that America had “goodwill” in other countries before Trump is laughable. Where? Latin America? Africa? In the Muslim world? We bombed the hell out of all those places long before Trump. This most recent Iran war has generated less outrage in the Muslim world than the war against Iraq 20 years ago. American foreign policy since the 1950s, fixated on fighting communism and then terrorism, has meddled with so…

The US is a big part of the customer base of the largest manufacturing economy in the world. China's economy blossomed via US and European consumers.

Re: The US is winning the AI race where it matters most: commercialization

#468
post #435
post #297

Earlier quoted context omitted.

Before AGI can choose for itself, it will depend on its creators to decide what it values and how it behaves. We can see how that works whenever grok gets the answer factual.

Am I the only one seeing the very obvious parallels to child rearing here...

No, it is one of the standard tropes in the field.

Re: The US is winning the AI race where it matters most: commercialization

#469
post #444

This makes no sense when you zoom-out. None of these companies, be it Anthropic, OpenAI, xAI, Google, Meta, Microsoft, are profitable in the AI department, they're all bleeding money and using funds their parent company and/or investors, primarily investors gave them. The Chinese models are keeping up with them, while offering the models for free and able to run on consumer grade hardware, and more importantly they t…

> None of these companies, be it Anthropic, OpenAI, xAI, Google, Meta, Microsoft, are profitable in the AI department, Citation needed. All reporting is that they are profitable on the inference side and all the VC money is going to building more data centers to run more inference. (Note that the coding subscription models are probably only break even on average - the money is in the API) > The Chinese models are kee…

What does profitable on inference mean? As far as I can tell, none of these companies have rigidly defined it, let alone it being a GAAP number. And yeah, if you subtract out all your R&D, payroll, sales, marketing, and other overhead, and get someone else to take on the debt or dig into their free cash flow to build the hugely expensive infrastructure on which you depend, it'd be pretty hard to not be "profitable". It's almost humorous how dumb of a metric "profitable on inference" is.

Ask yourself if AI was so profitable, why don't any of the big hyperscalers break out AI revenue in their earnings. OpenAI and Anthropic both project huge losses for the next couple years, it's not hard to find.

The real problem is, as the GP comment pointed out, that they can never stop training. As long as they're committed to building these behemoth models, the second they stop training, someone else will catch up and everybody will switch over because it's trivial to do so.

Re: The US is winning the AI race where it matters most: commercialization

#470

Earlier quoted context omitted.

Don’t take China on face value, they have every incentive to promote a grifting military industrial complex in the US while focusing on competing in manufacturing. An actual war would fix a lot of the grifting in the US as it would align interests. Pretending they’ll go to war over Taiwan and not doing it is an effective strategy for undermining the US.

I hope you are right, but unfortunately there is no particular reason to trust China's leadership anymore. They are not nearly as obvious at Trump, but they are not on a good path.

All China needs to do is do what they’re doing play the long game the United States is currently shooting itself in the head, if they’re smart, they should just sit back and watch the show by the year 2100 well you know. And coincidentally that also applies to Russia sit back and watch them do it to themselves.

In addition, some of the other countries like Canada, Mexico, Australia, and New Zealand had better get busy from within because they’ll be on their own. In the same applies probably to Europe.

Post reply on HN