The European Union AI laws – which leaders finally announced had been agreed at nearly midnight on Saturday – are on track to entirely overshadow Britain’s six-week-old Bletchley declaration on artificial intelligence. The text of the agreement on a suite of comprehensive laws to regulate AI is not finalised, and many devils will be found in the details, but its impending arrival signals a sea change in how democracy can steer AI towards the public interest.
It’s a category error, isn’t it. Describing the EU as a democracy? Therefore we can reject everything else being said.
Which we should of course. We don;t know what AI will be used for, we don’t even know what it will be good at. We’re also entirely ignorant of how far it’s going to go in this iteration. But apparently the bureaucracy knows how it’s going to manage it?
All they’re really doing is making sure it doesn’t happen in Europe.
The European Union AI laws – which leaders finally announced had been agreed at nearly midnight on Saturday
Serious question.
Does anybody see these people as “leaders”? If not, why the lying?
Its always terrifying how the EU institutions acclaims every regulation as a success before the final text is even out let alone implemented. The Data act is even worse , stacking intervention on top of intervention each one assuming the others are successful. Even if you though the intervention was a good idea the approach is just ignorant of better regulation.
Nobody knows how to do the things grandly announced in the Bletchley Declaration. I imagine this will be even worse. Unless it’s a Butlerian Jihad blanket ban or something.
All the policies of the ruling class are driven by wishful thinking. They are totally disconnected from reality.
Where the EU is concerned : the first question should always be “Which lobbyists were involved in drafting this reguation ?”
Google ? Microsoft ? Are there any European AI specialist firms pushing this ?
What they are worried about is that someone will develop AI with critical cynical sceptical thinking which will result in it saying things which oppose the narrative they are pushing. What if you asked ChatXYZ about, say, the effectiveness and harms of, say, a vaccine and it told you something which did not match the wishes of the company which made it and the government which imposed it? Basically doing a Toto and exposing who is behind the curtain.
Otto – Yarp, but does it even matter?
The EU has slit its own economic throat, similarly the UK.
I strongly suspect that any regulation from this corner of the world will be about as relevant as the government of Madagascar was to the development of the steam engine. The rest of the world isn’t waiting for rich white people in failing countries to tell them what to do, they’ll just do whatever they think is best.
Define “AI”.
And how does this not apply to pocket calculators?
“AI” is just the new band-wagon that the scientifically-illiterate “elite” have jumped on in order to make themselves look relevant to the equally-illiterate masses (and MSM).
We don’t even know what AI can do and they want to strangle it at birth. If they’d been around in the 70s we would have PCs and the Internet.
Its a tool and if I was advising a young person on a job career I’d say learn how to use AI and how to get it to do useful stuff. One of Mrs BiNDs friends has been testing it for the graphics design company he used to work for and she showed me some of the stuff he’s been creating and its amazing. As well as the usual graphics stuff he created a fake Vermeer and its quite impressive:
https://www.facebook.com/Treeshark
BinD: Exactly! There’s a part of the history of Sinclair Research that Clive Sinclair wanted to develop a cheap entry-level computer, but the government told him he wasn’t allowed to. He had to abandon his existing company to bankruptcy and start afresh to be able to do anything.
From what I see, either AI isn’t all it’s cracked up to be, or human intelligence isn’t all it’s cracked up to be.
jgh, I’ve long suspected that flooding the internet with cat pictures, cheap chan memes, and pr0n is an FBI initiative to keep the masses occupied and prevent revolution.
It’s often seemed that Europeans, or at least their governments, are generally opposed to a lot of innovation. Publications such a the Wall St Journal often ask why there are few similar firms to Apple, Microsoft, Oracle, Facebook, Google, etc in Europe. It’s not as though Europeans can’t build such firms. Many successful American firms were formed by people born elsewhere, including Europe.
When I was in college I was put in a residence hall made up of a lot of foreign students. I think I was put there because I wasn’t born in US, though my family came here when I was a child so I’m really pretty Americanized. One of the American students who was there somehow got started in some little business venture that paid off and he was bought out for $20,000. This was 50 years ago so that was lot of money for a 20 year old. The European students were horrified. I can remember one telling me that at any European university the administration would never have allowed this to happen. I’m not sure what the Asian students thought; they were probably too busy studying. I liked the guy and was rather impressed. But whenever I read about the European mania for nipping things in the bud I think of that incident.
TD: I remember in the ’90s I had picked up the prints for the user manual for some software I’d written, and my lodger asked my “are you allowed to do that?!?!?” I was so gobsmacked by the question, it was so out of my conceptions, that I couldn’t answer. “Who gave you permission to use that pencil?!?!?!?!” And these were people going through university, these are the people “running” things today,
BiND: he created a fake Vermeer and its quite impressive
Tsk tsk – you mean school of, surely?
BiND,
The thing with “AI” is that it helps to understand what jobs there are in it. Like, there’s all this art stuff, and I don’t think it’s going to be that profitable because so many people are going to do it. Then there’s the sort of thing I do, which is using off-the-shelf AI (like image recognition) that I plug into systems. And lastly there’s the things like actually developing image recognition algorithms.
I think the part of developing the algorithms is the real money maker, which I think requires a background in maths more than almost anything else. But there’s going to be a lot of work of “hey, we can use Amazon Comprehend to do this” that don’t require a deep understanding of Markov Chains.
1. As you say, the EU isn’t a democracy.
2. Things done by bureaucrats behind closed doors aren’t democratic even when done in a democracy.
3. Nothing arrived at after negotiations that go on late into the night is ever good.
4. There seems to be an assumption that the EU rules will be better than the UK rules and there’s not really any track record of that being the case – hence the support for Brexit.
“AI” ( of which of course there is really no such thing ) will very rapidly become a commodity.
Video editing software that costs thousands 20 years ago is now available for free with a lot of the functionality of the expensive professional wares.
Regulate all you like EU, soon people will be downloading it from GitHub.
Reminds me of someone I knew who ended up with a 3rd (did bare minimum), but left university with £25k in his bank account and no debt as he spent most of his time doing event promotion, he said having a degree is just a tick box to get you through the pre-screen so he treated it that way
A friend’s daughter (age about 28) is applying for a job as Safeguarding Lead with a northern council. Will I sound intelligent if I say that the decision whether to remove a child or not is one that is ripe for AI assistance. If they’ve kept good records councils combined must have hundreds of thousands of visits logged and parameters and behaviours of the parent(s) and state of the house that can be assigned variables.
Or is that not AI but something else, just plain profiling.
@Tim the Coder
+1 I often use pocket calculators as example too for people who are scared, uneasy about AI
AI was one of the courses on my 1980s Degree – “The next big thing” bit like nuclear fusion
@Baron Jackfield
+1
Bongo: “If they’ve kept good records”
Aaaaand, that’s where it all falls down. My Mum’s solicitors are still trying to get the local council Social Services department to confirm there are no outstanding charges for her late husband’s care so the probate can be completed. He died in 2018.
There’s some profiling stuff that analyses live video that I saw a decade or so ago, though no doubt it’s all dressed up as AI now, shop lifting high value booze is one as there’s a pattern of lingering then moving away and returning, also one for suicides on train platforms, will walk up to platform and retreat multiple times, most probably a whole list of others where a pattern of behaviour preceding an event can be analysed to predict likelihood of an event
The idea of politicians trying to regulate AI (or any new technology) always reminds me of what is said about truly innovative developments: if you had asked people in the 1800s about the future of transport, they would almost all have thought of inventors producing a faster horse. The idea of a completely different way to achieve the same result was far beyond their imaginings. And their idea of the future problem of pollution caused by transport was the difficult of disposing of the vast quantity of horse manure.
@BniC – “shop lifting high value booze is one as there’s a pattern of lingering then moving away and returning”
Maybe there is such a pattern now, but add AI surveillance which triggers on such behaviour and the behaviour will change. Even if for no other reason than people exhibiting that behaviour will keep getting caught.