Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Can Europe train a frontier AI model on the compute it owns?

github.com

Post reply on HN
Print view View original

261–270 of 313 posts

Re: Can Europe train a frontier AI model on the compute it owns?

#261

Post by mnewme »

sajithdilshan wrote: post #247
mnewme wrote: post #224

Earlier quoted context omitted.

Actually right now the reverse is happening. GDPR will become easier there is simplification coming, there will be less national goldplating

They have been discussing that for years, but still nothing happened. At the rate of the bureaucracy it would take few more years to even draft the laws to ease GDPR

You are not up to date, this is already in legislation process and needs the vote of the parlaiment, highly likely it will come. There is also a simplification package and a migration pact. A lot is happening right now

Re: Can Europe train a frontier AI model on the compute it owns?

#262

Post by garte »

aspenmartin wrote: post #216
FridgeSeal wrote: post #199

Earlier quoted context omitted.

LLM’s are nice. But if push came to shove, we could very easily continue on without them, or without them involved in every aspect. Being able to write cutesy SaaS apps faster is neat, but not like, essential, the way water is.

LLMs are a “nice to have” blows my mind. LLMs power autonomous drones that kill real people in real wars. Coding agents are the primary way people code. That and we get a GPT4->5 jump every 18 months. Are we on the same planet?

LLMs conduct strikes on children which a human may not have done. so wow, you got a frontier model but you lost something there in the process.

Re: Can Europe train a frontier AI model on the compute it owns?

#263

Post by bluegatty »

satvikpendem wrote: post #210
bluegatty wrote: post #134

Earlier quoted context omitted.

No it's not. Europe has no models to even block. The US has a guy who occasionally can screw things up for a few weeks, but who will be gone in a while. You have it upside down: the innovation and the stuff is the valuable thing, the laws are there to help us organize ourselves a bit after the fact. They're always a secondary concern to the extent that the vast majority of civilization is working with one another, do…

Exactly. Europe makes the process and bureaucracy the end itself rather than understanding that they are one part of a means to an end, of actual innovation. People don't call Europe a mausoleum for nothing.

Nobody calls Europe a mausoleum.

It's vibrant and productive, just not at the 'next level' and it lacks some industrial dynamism.

Re: Can Europe train a frontier AI model on the compute it owns?

#264

Post by moffkalast »

BjoernKW wrote: post #209
moffkalast wrote: post #82

Earlier quoted context omitted.

It's certainly a good question. On the idealistic side it's the right choice, people should have the right to have a say in their own data since it's implicitly copyrighted. GDPR has done wonders to prevent careless personal data leaks that are so common in the US, and other kinds of abuse. In a more practical view though I'm not sure if it'll do anything to stop job replacement from automation as such. Most corporat…

> GDPR has done wonders to prevent careless personal data leaks that are so common in the US, and other kinds of abuse. Has it? I still have to see evidence of that. What GDPR definitely has achieved, though, is people engaging in pointless busywork out of fear some busybody is trying to have them fined for being in violation of GDPR. > In a more practical view though I'm not sure if it'll do anything to stop job rep…

Hmm well the practical point of GDPR is twofold from where I see it: the process being too convoluted and risky turns off some corporations from storing data entirely (maybe it's not complicated enough yet in that regard), and the deletion requests where I can send anyone holding my data an email saying they have to delete it all. I do it sometimes, maybe it's just a placebo and half I secretly hope it is because once someone inevitably whistleblows it, they'll be appropriately fined in the billions like Google gets regularly.

> If a job can be automated that means humans ultimately can engage in more worthwhile endeavours

Yeah if that were actually the case. Seems like the plan is to just automate everything we can in broad strokes, then wait if anything happens to turns up to occupy that portion of the workforce. Nobody seems to have any idea what to do with vast amounts of unemployed people who aren't qualified to do anything anymore.

I think the very possible end result is that there won't be any immediate new jobs in a meaningful volume in the time span when they're needed. As you say, the timeline can be very short, and in the US it certainly will be. The idealistic future is that UBI gets implemented, automation gets taxed and redistributed, so the economy continues to work. But that's a fantasy with the current ring wing wave across the world where any kind of social service is seen as communist money burning for some reason, pocketing that extra wealth through corruption will be the priority. We've seen this again and again in countries where most of the income is dug from the ground in some form, which is economically the same as a few companies making it all through automation. The end result is usually not great for the population. With the AI Act being very anti-authoritarian, banning credits cores, facial recognition, etc. it's a step in the right direction to compensate.

The likely result is probably gonna be some kind of army service and an increase of international tensions to justify a draft when we realize there's only so many extra Wolt drivers a population needs, and then those will go to Starship too. With a war economy you can do pretty much anything to maintain stability, the numbers are made up and the protests don't matter.

If a delay in adoption can help bridge this intermediate gap without complete chaos, millions will suffer a lot less.

Re: Can Europe train a frontier AI model on the compute it owns?

#265

Post by thomascountz »

cyanydeez wrote: post #155
thomascountz wrote: post #96

Earlier quoted context omitted.

I'm not sure I follow how this metaphor plays out. Many stand to gain by harming and threatening Europe and many stand to suffer if Europe is harmed or threatened. Far be it from me to claim Europe a perfected system, but there are globalized interdependencies which exist here. I think that I'm not seeing the indicators of hard benefits which correlate to the hypotheses put forward by the financial markets. In which…

These AI models are powerful enough to be used as cyber threats, as a minimum. They may not prove out the other GDP or ROI claims, but the ability to both improve security and dismantle is means if you're not regularly building and improving these models on the software vulnerabilities your organization has, you are increasingly vulnerable to attacks. Since it's impractical to airgap your military and readinessness c…

Ok, now I understand better the metaphor. I don't understand your conclusion about AI hype, but at at rate, your earlier point is clearer.

The story I'm not readily conceiting to is that trillion-dollar frontier model training capex is the only thing that results in the defense capabilities countries need in relation to information security. From my understanding, the cyber wins being celebrated of late have come just as much (if not more) from how models are wielded than how they were trained.

Maybe the more honest framing of my question is: at what level of capability does sovereign training become the only viable path, and is that level the same as the trillion-dollar pretraining capacity TFA is actually asking about, or is it rather something more tactical/accessible? That is the part I'm genuinely more interested in, since the closer you are to benefiting from the commercial aspects of model training, the more likely it is that the answers will simply follow the money, so to speak.

Re: Can Europe train a frontier AI model on the compute it owns?

#266

Post by seb1204 »

nradov wrote: post #197
mnewme wrote: post #98

Earlier quoted context omitted.

European countries have higher freedom of press than the US. Bearing arms is not a human right in Europe, different culture. Europe has more human rights protections than the US and stronger enforcement of them, even against the state, by many metrics. Freedom of expression ends where other human rights begin, is protecting hate speech and Holocaust denial really something worthwhile?

So are you claiming that human rights are subjective and not universal? Or that it's acceptable for the EU to violate human rights if the USA is worse in some ways? Belarus is a European country. How is freedom of the press doing there?

It's doing fine

Re: Can Europe train a frontier AI model on the compute it owns?

#267

Post by ExoticPearTree »

adev_ wrote: post #254
ExoticPearTree wrote: post #117

Earlier quoted context omitted.

> isnt the eu basically a less federalist version of the usa? And what good does it do? The EU cannot speak in a single voice - there is no foreign minister, no defense minister, no whatever minister. Like the EU foreign function: you have a person claiming to be the EU voice, and then you have the foreign ministers of every EU member that can just say whatever they want if what the EU voice says is contrary to their…

> And what good does it do? The EU cannot speak in a single voice - there is no foreign minister, no defense minister, no whatever minister. Nor can Switzerland. And still it is one of the best country world wide both in term of living and economically. Distributed federal power like Switzerland trades quick decision making for resilience. If it might look up 'messy' on the surface, it is in fact a quality. A very va…

I've never seen the Swiss foreign minister being contradicted by another minister or by an official from a Canton.

But the EU foreign minister can be contradicted by basically any country president/prime minister or their foreign ministers if it says something that is not aligned with every other EU member.

Re: Can Europe train a frontier AI model on the compute it owns?

#268

Post by ExoticPearTree »

thinkingtoilet wrote: post #126
ExoticPearTree wrote: post #121

No, it cannot. The fundamental problem between the US and EU it is that in the US if something is not explicitly forbidden, it is allowed and in the EU if it is not explicitly allowed it is forbidden. You cannot have innovation at the speed and scale that you have in the US because the legislation is cumbersome and there is no unified market with the same rules.

It's not just that. The US models were trained partly due to massive theft. They illegally acquired (not used, that was legal) massive amounts of texts that they trained their model on. If you or me stole that much we would be in prison but if you're rich you get to do it and pay a small fine. It's not even the laws. It's in the US you're allowed to break the laws.

I am of the opinion that if it is on the internet, it is free. Just like I find free movies, so do AI companies find free books or whatever they need for their models. What's so hard to understand?

Re: Can Europe train a frontier AI model on the compute it owns?

#269

Post by ExoticPearTree »

okanat wrote: post #221
flopbob wrote: post #201

Earlier quoted context omitted.

It's sometimes very telling when other people think of you and your country. While I do agree we in Europe have sometimes too many rules, but I just can't understand the American obsession with freedom and being able to do things to the detriment of society

It's also quite weird that Americans are the first ones to go full judgemental mode as it is demonstrated in this comment section. When I visited there, I also found people to be overly reactive and judgemental of seemingly minor things and quite obsessed with wealth. I love the fact that nobody gives a flying fuck about anything other strangers are doing in Europe. That's not to say that there are no bureaucratic in…

> quite obsessed with wealth

You do know it is better to cry in a Mercedes than on a bike, no?

Re: Can Europe train a frontier AI model on the compute it owns?

#270

Post by lwhi »

JumpCrisscross wrote: post #78
rootlocus wrote: post #38

Earlier quoted context omitted.

We can debate the details and implementation but EU legislature is, at least in spirit, trying to protect human rights. What does it say about progress that the same laws that protect human rights also stifle innovation?

> EU legislature is, at least in spirit, trying to protect human rights. What does it say about progress that the same laws that protect human rights also stifle innovation? As you said, in spirit. In fact the EU’s AI Act is not really human rights legislation. (It exempts military and national-security uses.) Where it comes close, e.g. in seeking to ban facial recognition or social scoring, it does so clumsily. So i…

I don't think you're able to understand what the alternative looks like.

There is no way I want America's future.

Post reply on HN