This week, a grateful country learned from our Glorious Leader that the U.K. was to be not just one superpower, but two. At an event at University College London on Monday, Keir Starmer unveiled his plans for making Britain an “artificial intelligence superpower”. According to Reuters, this includes promises “to take a pro-innovation approach to regulation, make public data available to researchers and create zones for data centres”. The problem with this plan, however, is obvious to anyone who knows anything about AI and who has sufficient memory to recall Starmer’s other promise: to make Britain a “clean energy superpower” by 2030.
As has been widely discussed, Britain industrial energy prices are now among the highest in the developed world. This is a problem for AI because, unlike most computing tasks, it is energy intensive, better thought of as an industrial process. AI is not an application that runs easily on a laptop. And this is because AI does not typically run on normal microprocessors, such as the CPUs that sit at the centre of powerful desktop workstations, but on the graphics processors known as GPUs that enable 3D games, among other things. In such an application, the CPU pushes the work of rendering graphics to the GPU, leaving the CPU free for other, less intense tasks.
To read the rest of this article, you need to donate at least £5/month or £50/year to the Daily Sceptic, then create an account on this website. The easiest way to create an account after you’ve made a donation is to click on the ‘Log In’ button on the main menu bar, click ‘Register’ underneath the sign-in box, then create an account, making sure you enter the same email address as the one you used when making a donation. Once you’re logged in, you can then read all our paywalled content, including this article. Being a donor will also entitle you to comment below the line, discuss articles with our contributors and editors in a members-only Discord forum and access the premium content in the Sceptic, our weekly podcast. A one-off donation of at least £5 will also entitle you to the same benefits for one month. You can donate here.
There are more details about how to create an account, and a number of things you can try if you’re already a donor – and have an account – but cannot access the above perks on our Premium page.
To join in with the discussion please make a donation to The Daily Sceptic.
Profanity and abuse will be removed and may lead to a permanent ban.
There is no post industrial economy. We are in an industrial economy and always will be
Ecolunacy has an anti-human streak to it. For them humans seem to be at the centre of all destruction and ecological degradation.
However, when you compare human intelligence to artificial intelligence, we are the very epitome of efficiency and sustainability (I hate that word really).
With a bit of organic material each day the human supercomputer can function without any problem. AI in contrast needs vast amounts of energy to perform a miniscule subset of functions that humans can perform.
AI can perhaps do much much more quickly that handful of things that humans can do more slowly, but it needs insane amounts of energy to perform it.
DNA is really a wonder. We should appreciate it more.
Reminds me of this: Computers that power self-driving cars could be a huge driver of global carbon emissions | MIT News | Massachusetts Institute of Technology
“In the future, the energy needed to run the powerful computers on board a global fleet of autonomous vehicles could generate as many greenhouse gas emissions as all the data centers in the world today.”
“The researchers also found that in over 90 percent of modeled scenarios, to keep autonomous vehicle emissions from zooming past current data center emissions, each vehicle must use less than 1.2 kilowatts of power for computing, which would require more efficient hardware.”
“When they used the probabilistic model to explore different scenarios, Sudhakar was surprised by how quickly the algorithms’ workload added up.
For example, if an autonomous vehicle has 10 deep neural networks processing images from 10 cameras, and that vehicle drives for one hour a day, it will make 21.6 million inferences each day. One billion vehicles would make 21.6 quadrillion inferences. To put that into perspective, all of Facebook’s data centers worldwide make a few trillion inferences each day (1 quadrillion is 1,000 trillion).”
The few people that have bothered to do the maths (basically the architects of AI) are not surprisingly pushing hard to get lots of new nuclear power stations.
I don’t know if you saw that somewhat bizarre presentation Elon Musk did of his AI robots a few months ago. He arrived in a supposedly autonomous vehicle and then the robots shuffled out to “mingle” with the assembled guests.
To me the whole thing reeked of smoke and mirrors. The so called autonomous vehicle came along empty roads in this isolated Hollywood like studio setting.
And then when the robots came out, to me it just looked like ChatGPT on legs. In other words, nothing new.
What struck me was the staging of it all. It was at night, dark, with very deliberarte lighting. It looked to me like a David Copperfield show.
Since then, most of the news I read about AI I assess very carefully to see how much is tangible news and how much is hype (may, should, could, will statements). And most of it to be honest, seen through that filter seems like hype to me.
And so when I see these tech uber-bros talking about AI, what I see is guys hyping it up to lure enough investors long enough so that they can parachute out with billions before anyone discovers there is no way to make money out of it.
I think driverless cars is a pretty dumb idea, solving a non existent problem.
I didn’t see the presentation but can imagine it.
I wish I understood more about what AI might be capable of in the future and what it is currently good for. We use it at work (software) for code completion, hints and suggestions. My guesstimate is that it adds a few % to the productivity of the team. We may do a bit more R&D to see if a locally deployed Large Language Model could produce a bigger uplift that would significantly enhance our capability. So far I have not found much information on real world major projects that have brought success – either there have not been many or people are not sharing the information publicly. My suspicion is that we can do more with it but it will not be revolutionary and that I will still need my whole team and I will need them to use the tools with intelligence. From what I can see, the skills required to use it well as the same as those required to develop good software. I think currently it will be a case of GIGO. Of course this is in the realm of software which probably lends itself to this kind of stuff. Using it more fully would require us to make a significant investment in hardware which may not be viable, or use a cloud platform that enables us to share the underlying hardware with lots of other people – but this is probably going to limit a solution to something we use as and when, not something that is going full steam ahead 24-7.
In short, probably overhyped right now but I simply don’t know where it might lead. My guess is that for it to do anything amazing we’d need a revolution in hardware along the lines of quantum computing, which like nuclear fusion always seems 10 years away from being useful.
Driverless cars mean fewer paki taxidrivers – don’t knock it.
My experience with AI is that it does some secretarial work quite well, like writing first drafts of things.
I’ve also heard it can be quite useful for producing images.
And like any tool, it’s usefulness depends largely on the skill of the user, the precision of the instructions in this case. I imagine that’s the experience using it in computer programming?
But the big question is, how much does it cost. Actually cost, not what it’s currently priced at, but what it really costs. My sense is that it’s not cost effective given the energy it requires, but it’s one of those type of projects that they hope it’ll work out somehow if they keep going.
But I wouldn’t put a whole lot of value on this opinion. it’s really based on slivers of information.
I would go along with everything you have written – it mirrors what I have observed. My friend who is training to be an illustrator is worried because of the image generation. I don’t like the writing it produces much.
I watched a YouTube video from a bloke who downloaded various LLMs and tested them – tiny ones ran OK on sensible hardware, the larger ones ran like a dog on some stupidly powerful expensive thing that Dell had lent him, which looked like something out of Star Trek. The larger the LLM the “better” the responses are supposed to be. So yes my guess is that the LLMs in the cloud are a huge loss leader and that if they were charging cost price then my illustrator friend’s job would be safe.
I wonder if they could be sued for anti-competitive practices.
I don’t know exactly what legislation there is around that but I have a vague sense that there are anti-dumping rules, certainly against unfair overseas competition.
But you’d think that is one useful piece of legislation they could put in place to protect against the decimation of employment. A bunch of deep pocketed billionaires should not be able undercut human labour with massive.loss making AI.
Possibly. I guess they would argue they are doing humanity a favour.
Anyone who thinks that AI can replace a human illustrator is an idiot. Maybe sometimes what the client wants is just perfect and AI produces something decent – but although the customer is always in the end right, the role of a supplier of creative services is to consider what they think would work best and make suggestions based on their experience and ingenuity. I don’t think AI can do that.
Emperor’s New Clothes you mean?
Yet another gov’t initiative that has the banner of “lead the world” and requires “picking winners”. Neither of these goals have worked out by this or any gov’t, probably anywhere on the planet.
Starmer hasn’t got a clue.
The UK becoming an ”AI superpower” – it’s all bollox. It’s just a soundbite, a trendy, “cool”, fashionable phrase he picked up and uses to create an illusion that he’s got some ideas. He hasn’t.
The guy is an empty suit.
Exactly, it’s just something that’s come out of his mouth without engaging any sort of critical thinking along the way. Maybe he heard someone else say it, who knows. I’m sure he doesn’t have a clue what it means, and nor do I, but then I don’t pretend to be capable of running the country. And I’ve said it so many times now – if we really WERE facing climate catastrophe, if it really WAS two seconds to midnight (or at least if Starmer believed it) then AI would be the very last thing on his mind, utterly unnecessary as it is to human survival…
I’m not sure anyone is “capable of running a country” except God. They could start and stop with “first do no harm”. We must stop thinking of these people as some kind of ship’s captain.
He’s like the scarecrow, tin man and cowardly lion rolled into one.
It is a tarted up version of Cameron and Osborne’s genius wheeze, HS2, that would transform the economy by getting businessmen from London to Birmingham as quickly as possible.
Rolls Royce are developing Small Modular Reactors, so the UK could be a super-power in AI and Climate … except I don’t think the government likes nuclear, either.
The problem with Small Modular Reactors is that they might, just might solve a problem that Kneel does not want solving ie power outages.
When people have no power and therefore no heat and light they fall evermore under state control – mobile phones switched off, no power to cook with, cold homes, no stupid telly box. This doesn’t appeal to the masses but if government can avert these problems for say 50% of the time then government appears amenable even when it caused the problems in the first place.
Net Zero fallacy and folly direct consequence of worshipping false god of computational modelling, latterly joined in devilry by other false idol of artificial intelligence.
Cue once more the Feynman lectures of 60 years ago, just as mainframe computers began to proliferate around the planet:
https://www.youtube.com/watch?v=EYPapE-3FRw
“If it disagrees with experiment, it’s wrong. In that simple statement is the key to science. It doesn’t make any difference how beautiful your guess is, it doesn’t make any difference how smart you are, who made the guess, or what his name is…
…If It disagrees with experiment, it’s wrong. That’s all there is to it.”
Someone needs to sit this man down in the corner of a classroom on a small chair and explain to him that he can’t be Batman and Spiderman at the same time.
Amazing the lengths the man will go to, to avoid an inquiry into child grooming gangs, something to hide maybe?
Such childish language isn’t it? Do they still really feel that they are in the infants playground?
Britain is only a world leader in stupidity, vanity projects and wishful thinking! And Wokeness.
How can you be a leader in anything if your policies are destroying the economy and the energy sector? By design!
You can only be a leader if you have followers and the brain dead zombies that make up the PLP DON’T count. Nor do the non Lib Dems.
Lib DIMS.
Please.
This has to be a joke right? What – you mean he’s serious? Forgive me, this is lunacy. Soundbites smacks of desperation.
Ah, but the Glorious Leader has already made his country a Child Grooming Superpower. What more is necessary?
AI, heat pumps, electric cars, green hydrogen. That wind and solar is going to have to do a lot of work.
Do politicians “work things out”? I mean they work out whether they can get elected (though these days they don’t seem to care much about that either) and misbehave in office without getting caught, but done seriously think they sit down for days and weeks on end with piles of information relevant to an issue and try to come to an honest, logical conclusion without prejudice? Seems like a fantasy to me.
Yeah, complete fantasy.
But they are just salesmen whose primary skill is sniffing out what the public is prepared to buy.
The best possible outcome for any of us is that the “solutions” that are shoved in front of their noses are the best ones for us and that by some amazing miracle they happen to also be aligned with their own interest (basically their careers).
The chances aren’t good, but maybe from time to time..every few years perhaps… on some minor issue….
I think the best we can hope for is that the “solutions” are not catastrophically damaging and expensive. They can’t even do the basics like protecting our borders.
Starmer’s government is always going to be a British bullshit superpower, telling whatever fibs it can come up with in order to distract people from the fact that it’s simply wasting their money on itself and its cronies (like Dale Vince). Self-proclaimed AI developers would love government gazillions just as much as self-proclaimed green entrepreneurs: Neither of both are ever going to produce anything useful. Hence, government must provide the earnings the market wouldn’t (for long).
Thanks for digging out the driverless cars story. That ground to a halt the moment these vehicles had killed enough people that the question of legal responsibility arose in earnest. And if there’s something no software manufacturer is ever going to do, it’s to accept legal responsibility for the performance of its products. That’s tacitly accepted when the outcome is f****ed up Word documents or malfunctioning coffee machines or electronic payment devices. But tacit acceptance of dead people as price for “progress” didn’t manifest itself. Hence, no more “driverless cars”.
I was looking forward to making Demolition Man a reality!
In other news, Two Tier Never Here will be facing his first byelection as the Labour MP who attacked a member of the public has pleaded guilty and been sacked by Labour. He will no doubt be happy for the process of removing him and the byelection held to be as slow as possible in the vain hope his Student Union government will improve.
And yet in reality innovation and economic growth are encouraged by getting ‘polict’ out of the way. Cut red tape, reduce the need for red ink.
The big joke is, of course, that this slimy excuse for a lawyer spent months telling us how serious he is and how the adults were going to take over and remove the snake oil salesman, only to come down the road with wagon loads of the bloody stuff trying to convince us that what they’re flogging is a so much improved version of the snake oil, so improved that we shouldn’t even call it snake oil.
UK to be a World Leader in Net Zero Intelligence…
Let’s not forget that in the HitchHiker’s Guide to the Galaxy, the plutonium rock band Disaster Area overcame the objections of environmentalists to their huge gig on one planet by having them all shot.
I commend this idea to the house.
I must re-read it. I’d forgotten.
I’m sure we’re missing that the UK is going to have AI during the day and when the wind is blowing…. It’s why he asked regulators for growth ideas… Planning his way to nowhere
Let us hope there is some real intelligence programming the AI computers. Judging by the Post office Fujitsu computer and the Fujitsu MyOracle system recently installed at Norfolk County Council this is unlikely