The EU's AI Act is now in force
techcrunch.com
techcrunch.com
My take is that this act is an attempt by the EU to build some AI companies. They failed at the cloud - where China massively succeeded. They do not want to make the same mistake again.
There are provisions to (1) cause trouble for entrants (certifying LLMs), (2) provide national infra for testing (subsidize).
The EU also view it as an instrument of soft power as many countries end up basing their own law on the EU as they are often first mover.
> I also think greenhouse gases are a global issue, not an EU issue - please leave politics out of it.
Like it or not, the EU has regulations to limit its GHC emissions and did pledge to reach net zero in 2050 something it is not doing yet and is steering through regulations.
Please leave politics out of... global warming?
I mean, that seems like a pretty tall order.
by kicking western companies out
"national security laws that prevent non-Chinese companies from owning certain critical infrastructure assets"
https://www.datacenterdynamics.com/en/news/aws-sells-assets-...
What exactly does this mean? Will AI companies allow copyright holders to opt-out of having their data or works used as part of training? Are there tools to prune data from already trained models? I'm not sure how questions related to copyright are going to be resolved with existing AI models.
Despite the strong lobbying from US companies echoed here ("but think about our poor shareholders"), the regulation is fairly raisonnable, generally in the public interest and it’s very much possible to develop software while respecting it.
This is Europe not the US. I find it hard to believe that after years of GDPR Americans still fail to understand the way European courts work.
The worst that will happen at first is a small fine and you are more likely to be hit with a warning and an explanation of what you should have done. Fines only ratchet up from there if you intentionally stay non compliant.
Downvote as much as you want. I am eagerly awaiting all your exemples of insane fines levied on small companies for GDPR non compliance. You will be hard pressed to find some. Large companies with legal teams which are intentionally infringing however are hit as they should be. Strangely they are also the one beating the drum the loudest against these regulations, go figure.
90% of the complaints you hear about here are more on the basis of “how dare they do that to our American companies” that actually based on the law anyway.
> To whom does the AI Act apply? The legal framework will apply to both public and private actors inside and outside the EU as long as the AI system is placed on the Union market, or its use has an impact on people located in the EU.
https://ec.europa.eu/commission/presscorner/detail/en/qanda_...
> any society that will give up a little liberty to gain a little security deserves neither and will lose both
But Franklin lived in a pre-industrial society. He saw two sides in the balance: citizens vs government. Corporations, such as they were to that date, were in check to the throne.
Here's a reformulation for the internet age:
> Any society that will not give up any liberty to the government will instead give up liberty to people like Zuckerberg
I still argue that a traditional libertarian attitude - by which I mean viewing the balance between govt and citizenry as a zero-sum game - as leaving the flank wide open. It's a three-body problem, and government and corporations are closer natural allies than the citizenry is with either.
I read that as a 'it depends'. If, e.g. that company develops an 'AI financial adviser', which is later found to conspire to stock manipulation (pump&dump or otherwise) it might very well have an impact on people located in the EU, even if sold and used only outside.
> no recording in public
this is the only one that sounds like good common sense but isn't even regulation it's just law any sort of legal system would enforce.
all these "safety" things means ai offerings are all useless like google after 2003 when anything you query like "list of fad diet" is transformed into instead of an answer to that "problematic" question, an answer to "NVM LOL HERE IS A LIST OF HEALTHY DIETS"
> can't train on copyright material
that's so dumb. this world has no interesting copyright content coming out for 2 decades, it's literally all sellouts. losing to AI (which will happen regardless of whether the training material contains any copyright content) is exactly what they deserve. don't forget that these are also people who sue random civilians for "damages" that only exit in their head. sellout culture no longer even applies to musicians and book authors, it's also the majority of software too, with people just creating software for part of their resume, or some crowdsource shit, or a blog. any software anyone makes in a corporation is just to host an ad or be a dev tool for people who just host an ad
EU regulations are law.
> any sort of legal system would enforce
Many countries (the US, for instance) do not have such a blanket ban. For instance, see this nonsense: https://www.independent.co.uk/tech/friend-ai-artificial-inte... - its whole value proposition is indiscriminate public recording.
https://time.com/6288245/openai-eu-lobbying-ai-act/
>Still, OpenAI’s lobbying effort appears to have been a success: the final draft of the Act approved by E.U. lawmakers did not contain wording present in earlier drafts suggesting that general purpose AI systems should be considered inherently high risk. Instead, the agreed law called for providers of so-called “foundation models,” or powerful AI systems trained on large quantities of data, to comply with a smaller handful of requirements including preventing the generation of illegal content, disclosing whether a system was trained on copyrighted material, and carrying out risk assessments. OpenAI supported the late introduction of “foundation models” as a separate category in the Act, a company spokesperson told TIME.
Given how much we see these 'general' systems in political propaganda (think of all the Twitter far right propaganda unmasked by users replying with 'ignore previous instructions') I'm not sure I agree that these models do not fall under the high risk label.
The rules for high-risk systems are understandable, mostly increased human oversight and increased transparency: https://artificialintelligenceact.eu/section/3-2/
I understand that OpenAI does not want increased transparency in line with being the least transparent AI company out there. It's not like they don't have the money to follow these rules.
IMO, political propaganda is still low-risk; it simply means that citizens are panzying around and believing anything that pops in their feeds, and the correct counter-measure is to ban or regulate social media or to find a mechanism to encourage citizens to make better judgements (say, a tax deduction after passing a maths and modelling exam, a check for number of well-researched books read, or anything better than my 20-seconds-of-consideration rant).
I also like the current classification, but I hope that as AI's get better and more dangerous, our regulations will evolve accordingly.