World leaders ink AI safety pacts while Musk and Sunak engage in awkward bromance
- Reference: 1699017315
- News link: https://www.theregister.co.uk/2023/11/03/ai_summit_safety_pacts/
- Source link:
With various types of AI being routinely built into consumer and business-facing software, developers agreed that governments have a role in overseeing external safety testing of AI models. The move is designed to keep track of frontier AI models sitting solely within corporations.
Australia, Canada, the EU, France, Germany, Italy, Japan, and South Korea have all signed up to the agreement, although China – which is making huge investments in AI – was not a signatory, although it did attend this [1]week's UK AI Safety Summit during which the agreement was made.
[2]
The governments also declared a "shared ambition" to invest in public sector capacity for testing and other safety research, and to share the results with other nations. The idea is to lay the groundwork for future international progress on AI safety in years to come, the UK government said.
[3]
[4]
The AI testing agreement, which carries no legal weight, followed the [5]Bletchley Declaration , named after the home of the UK's pioneering Second World War code-breaking mission. Twenty-eight countries and the EU signed up to the mission statement, aimed at setting the agenda for addressing AI risk. Included was a desire to identify AI safety issues of concern and build a "shared scientific and evidence-based understanding of these risks."
[6]Yeah, that oughta do the trick, Joe... Biden hopes to tackle AI safety with exec order
[7]AI luminaries call for urgent regulation to head off future threats, but Meta's brainbox boss disagrees
[8]AI safety guardrails easily thwarted, security study finds
[9]Japanese PM says international AI regulations will be here by Christmas
Bringing to a close the UK's AI Safety Summit yesterday was the unedifying spectacle of Prime Minister Rishi Sunak interviewing tech entrepreneur Elon Musk, the world's richest man.
Beginning with nauseating sycophancy, Sunak read Musk a Bill Gates quote, saying there was no one in our time who had done more to push the boundaries of science and innovation than the Tesla CEO. It's perhaps a shame Musk [10]doesn't seem to share the same respect for Gates . Perhaps not.
The discussion then dragged through future AIs needing an off switch, and so on, but overall, Musk said AI would be a good thing, providing friendship and "a future of abundance, where there is no scarcity of goods and services."
[11]
Sky News TV journalist Sam Coates declared "it was all just mad," involving 40 minutes of "softball questions" in which Musk agreed to take questions from business leaders, but not journalists.
Coates said Sunak didn't see Musk as a political force prone to voicing views on Ukraine war and Gaza/Israel, with interests in the politically powerful Starlink satellite system, and failed to challenge him. Instead, Sunak focused on "selling Britain to this guy who is already the richest man in the world" in the hope Musk's celebrity sheen might rub off on him. ®
Get our [12]Tech Resources
[1] https://www.theregister.com/2023/11/01/uk_ai_summit/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2ZUUnO@jkqqBCo6XHvIC2igAAANU&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZUUnO@jkqqBCo6XHvIC2igAAANU&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33ZUUnO@jkqqBCo6XHvIC2igAAANU&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://www.theregister.com/2023/11/01/uk_ai_summit/
[6] https://www.theregister.com/2023/10/30/us_president_to_sign_new/
[7] https://www.theregister.com/2023/10/24/ai_safety_regulation_open_letter/
[8] https://www.theregister.com/2023/10/12/chatbot_defenses_dissolve/
[9] https://www.theregister.com/2023/10/10/g7_ai_rules_coming/
[10] https://www.theregister.com/2023/09/11/gates_musk_beef_biography/
[11] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ZUUnO@jkqqBCo6XHvIC2igAAANU&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[12] https://whitepapers.theregister.com/
Oh, we've got Hoverboards. They just don't hover. And have been banned from the rail network.
(Are those things still on sale? I haven't seen them in a while...)
Correction. Hover boards that actually work (like in Back To The Future) and flying cars that actually work (like in Blade Runner)
Re: Timescales?
One day, perhaps. Looking at office work, and administration tasks in general, most will quite possibly become fully automated within a generation or three. Looking at manufacturing, robots can already do a lot of the repetitive or predictable work, I think it might take a bit longer to get all tasks under AI and robotic control (though we'll probably get a long way down that road fairly quickly in the more developed economies). Agriculture/farming is already fairly mechanised and automation is quickly gaining ground (again, in the more developed economies).
There are significant caveats, though.
Who will make decisions? How will it all be governed? Where will any human oversight fit in? Those are all work to some people, so the "nobody needing to work" criterion has to address even that.
Once there is limited work, how will the economy be managed? If people aren't needed for work, what will they be needed for? Even leisure activities (including entertainment) will need to become autonomous? If nobody earns a wage, the state will need to support.
Who will invest in the poorer economies to support those who fully depend on being able to work for a wage? For example, automating clothing "sweat shops" whilst concurrently giving the workforce wages without them having to do anything for them.
And a lot else I've not even considered...
The ultimate goal may be achievable but the current economy and (global and national) government is unlikely to be able to manage the transition.
Musk's statement is, to me, akin to predicting that the Sun will eventually die - and the timescale needn't be all that different!
Re: Timescales?
Looking at office work, and administration tasks in general, most will quite possibly become fully automated within a generation or three.
You can already observe quite funny thing at some offices. Workers send emails generated by ChatGPT to their colleagues, then their colleagues put these emails to ChatGPT with some context and reply back. Rinse and repeat.
Remember the 70's promise's
When I was growing up, we were promised that robotics would free us from drudgery, and by the time I became an adult, we would all live a life of leisure while our robotics slaves did the work for us.
The reality was that it increased the gap between the have and have nots, forcing people into jobs that were not worth investing in Robotics, while the profits went to an increasingly small mega billionaires, who used there cash to wield soft power to protect their status
I have no confidence, AI is not going down the same route
Re: Remember the 70's promise's
The worrying thing is that these billionaires always seen to want to force us to do things "for our own good" and shape the world the way they think it should be. Gates, Musk, Soros, Thiel and Schwab and ALL megalomaniacs* and are all very powerful. And they all think they are doing good. And they are all technocrats. And we can't vote them out, even though they have much more power than, say, the PM.
*just a few random examples, there are many more who fly into these conferences in their private jets.
A classic pairing
Of reality detached morons (musk and sunak)
Re: A classic pairing
This whole thing does my head in.
What the fuck is the prime minister doing there? Regardless of how useless you may think he is (and I do), surely he has more important things to attend to? As for Musk - a man trying to convert Twitter into a unified-everything app we should trust with our money, after he has just lost fifteen-billion dollars due largely to mismanagement - words fail me.
They're a couple of punchlines to a not funny joke - both trying desperately to divert our attention away from the fact they have failed at their actual jobs .
Re: A classic pairing
Sunak is grasping at straws.
Today I saw an electronic billboard with Sunak and Musk with some headline about "landmark AI something" and on the left there was a woman with a child sitting on a cardboard and eating food with their bare hands and then there was a guy harassing drivers in stationary traffic begging for money.
That was quite surreal, but shows how Sunak and Conservatives are detached from reality.
Musk isn't invited back. Ever.
Seems team Rish! are apoplectic that his Muskness suggested a job-free future. It's hard to imagine something so opposed to their plans for the UK.
However seeing Rish! squirm was worth it.
And generally I am no fan of Musk.
Re: Musk isn't invited back. Ever.
Aren't jobs for plebs?
Re: a job-free future
Without supporting His Muskiness in any way, shape or form, one has to admit that the end result of automation, AI and robotics means that, one fine day, the only humans who work are the ones who will actively want to. Everything that needs doing will be accomplished by automatons, with defined procedures that have been refined over decades, then centuries.
I refer to [1]The Expanse for a clear idea on how this could work : citizens, upon graduation, would decide if they wish to have a life defined by Basic (revenue, support, health care, etc.. I guess), or if they want to actually work for their subsistance and attain a career worth something.
That is one idea. I suppose there are a number of possible variations. But the point is : we're doing everything we can to work as little as possible. One day, that will mean that we don't have to work anymore.
So we might as well talk about it, I guess. Even though I would really prefer that a failure like Musk not be the one to champion the idea.
[1] https://en.wikipedia.org/wiki/The_Expanse_(novel_series)
Re: a job-free future
we're doing everything we can to work as little as possible. One day, that will mean that we don't have to work anymore.
And yet we work more than ever before.
The idea that AI somehow replace humans is nonsense. Humans always adapt and there will be always something to do.
Outcomes?
" The AI testing agreement, which carries no legal weight " followed " Twenty-eight countries and the EU [signing] up to the mission statement, aimed at setting the agenda for addressing AI risk ". Merely 'aimed at'?
So nothing binding on anyone has actually been achieved. Rather like climate change conferences really (and party conferences and ...). It sems we're in an age where aspirational waffle is the primary national (and international) product, so maybe ChatGPT is the right way to go -- automated waffle generation without the need to waggle the jaw.
Capture
Australia, Canada, the EU, France, Germany, Italy, Japan, and South Korea have all signed up to the agreement, although China – which is making huge investments in AI
This is all about regulatory capture and "hot air" Sunak, who is on the way out, is likely looking to secure the market for big corporations and keep potential competition at bay, hoping, in my opinion, that it will land him a nice and lucrative job somewhere in the bay area to stroke his fragile ego. After all he is still being seen as relatively poor husband of his rich wife and he is desperately trying to be his own man.
But I digress. You can see how much of a hot air it is, because China has not signed it. Why would they put shackles on their own economy?
Probably they had a good laugh at these neo-liberal idiots.
Timescales?
It's all very well predicting that we'll be living in The Culture (which is where Musk gets many of his ideas, including Neural Lace) and we won't need jobs...
but he's put no timescales on this. It's almost certainly going to happen in less than 1,000 years. In the lifetime of Sunak's government? Nope. My life? very unlikely. My grandchildren's lifetime? Could be, but still pretty unlikely.
I'm still waiting for flying cars, hoverboards or, I dunno, free public transport.
I do think AI is pretty cool, and I'm hoping that soon I'll be able to say "Computer, take my last year's expenses spreadsheet and copy everything to this year and update the dates and amounts for me" and it will go off and dig into my bank accounts and figure everything out.
Right now, I feel that companies are trying to drum up a lot of hype about the future of AI, so that they can get a big slice of that lovely VC funding. More bloody vapourware.
Putting in a "framework" for this would be like the Wright Brothers asking for an oversight board, to consider the potential impacts of plane engines, or something (please think of a better analogy, it's Friday and I'm tired).