And here is where Britain’s particular brand of suicidal virtue-signalling becomes lethal. The Liberal West, and Britain most zealously, has spent fifteen years chasing Net Zero with the fervour of a medieval flagellant. We’ve shuttered coal, dithered on nuclear, blanketed the countryside with unreliable windmills, and now face the grim prospect of energy rationing. The National Grid’s own forecasts admit that data centres alone could consume 7-10% of UK electricity by 2030, and that’s before the real AI boom hits. Microsoft, Google, and the rest are already scrambling for power purchase agreements that dwarf entire cities. Yet our political class still preens about “green leadership” while quietly preparing the public for blackouts and sky-high bills.
|
|||||
|
We are developing the social individualist meta-context for the future. From the very serious to the extremely frivolous... lets see what is on the mind of the Samizdata people. Samizdata, derived from Samizdat /n. - a system of clandestine publication of banned literature in the USSR [Russ.,= self-publishing house] Authors
Arts, Tech & CultureCivil Liberties
Commentary
EconomicsPolitics |
Samizdata quote of the day – the coming stormFebruary 12th, 2026 |
 85 comments to Samizdata quote of the day – the coming stormLeave a Reply |
Who Are We?The Samizdata people are a bunch of sinister and heavily armed globalist illuminati who seek to infect the entire world with the values of personal liberty and several property. Amongst our many crimes is a sense of humour and the intermittent use of British spelling. We are also a varied group made up of social individualists, classical liberals, whigs, libertarians, extropians, futurists, ‘Porcupines’, Karl Popper fetishists, recovering neo-conservatives, crazed Ayn Rand worshipers, over-caffeinated Virginia Postrel devotees, witty Frédéric Bastiat wannabes, cypherpunks, minarchists, kritarchists and wild-eyed anarcho-capitalists from Britain, North America, Australia and Europe. CategoriesArchivesFeed This PageLink Icons   |
|||
All content on this website (including text, photographs, audio files, and any other original works), unless otherwise noted, is licensed under a Creative Commons License. |
|||||
So true. And so unnecessary. CO2 has never been an important driver of climate and never will be. The modest radiative effect tends towards saturation, so each addition of CO2 has less impact than previous ones. The whole climate panic is based upon wilful misinterpretation of the science.
Just waiting for “The Day the Lights Went Out” kinda moment in the UK. Sure, we had the miners strike, but that was at least coordinated. It still took out the Heath Government of 1974.
This time around I’m expecting a balancing failure leading to a systemic failure. Given the malinvestment in windmills and solar panels instead of nuclear, it’s inevitable.
I don’t just blame Labour. The Tories need to take their share of blame for enabling this farce as well.
Regarding data centers and the AI boom, there is no reason for those to be located in Britain, they can be located overseas or in orbit. So if Britain makes it impossible or overly expensive to build them, then Britain will, simply speaking, sit this one out.
Which is a problem because “this one” is about to take over a large portion of the world’s economy making Britain even more of a beggar state, with utter suzerainty to the USA.
Fraser:
Agreed. If British electricity is expensive and unreliable, there would be no case to build AI centres here, unless the government announces massive subsidies. Subsidies are the only reason we have renewable energy, so it would make sense to waste more money on paying for AI centres to be built here. This is “investment”, and therefore a good thing.
Fraser,
Look at this.
Ha! California beat you to it! We win!
@NickM
Look at this.
Not sure the point you are making Nick (though thanks for the interesting link.) If your point is that electricity is expensive here too, my honest opinion is that the majority of data centers will be in orbit not on earth, where the cost of electricity is purely capital and not marginal. And of course that means that SpaceX will utterly dominate that market place. So perhaps the aforementioned suzerainty will be to Musk and not the United States.
FWIW, there will, for sure be data centers terrestrially for some specific niches such as bulk training an low latency responses. But the majority of cycles will be orbital, IMHO. And, FWIW, these cycles will not be spent on answering dumb questions on Grok or creating pictures of naked celebrities (though I’m sure there will be plenty of that), but on transforming the very ideas underlying the economics of he world.
BTW, I was listening to Musk and he said something that I thought was quite jaw dropping in a casual aside. If you are not a programmer it might not mean much to you, but for us programmers it is just a WTF moment. As a brief overview for the non programmers — the way software is written today is that the programmer creates it in a high level language like Python, C# or Rust that is designed to make the program easy for a human to understand. Then a tool called a compiler transforms that code into a set of binary instructions that the computer cpu or gpu understands and can execute. For example you might say something like: tax = (income – deductions) * tax_rate; which would be transformed about a dozen bytes of data. Maybe looking something like this (just making these numbers up): 6cbb 0113 94e7 4bbc a9ab b9cb ed79 ae49 e5c0 7f53 5aa7 4519 b726 d266 175e 1efd 60d6 50bf 47c7 4d45 9772 03c4 574f 5475.
What Musk said is that the AI coding tools xAI is building will entirely skip the high level language and generated the binary instructions directly. Why is that shocking? It means that human programmers simply will not be able to understand the code that the AI is generating. Certainly not at any macro level. And, based on what Musk (and his usually utterly unreliable timeline predictions) said, is that they will be doing this by the end of the year.
Which is to say these tools will mostly cut humans out of the loop. Which I think is jaw dropping.
If we can’t stop the politicians listening to the ‘scientists’ is there some way to make them listen to the engineers? “You realize how many raw materials will be needed, and how much maintenance will be required?” Musk is definitely an engineer.
@Ellen
If we can’t stop the politicians listening to the ‘scientists’
Politicians listen to the people with the opinions that justify policies that advance their wealth and power. I doubt too many politicians really deeply believe all that climate crap or for that matter the Covid stuff, rather they use it as a pretext for advancing their agendas. Do left wing American politicians really believe that we will be better off without the police? Of course not, they are usually surrounded by armed guards, but it sure does help them get elected and raise money.
I was going to write something about the nature of belief here, and what it actually means to believe something, but that sounds really boring, so I’ll skip for now.
The madness of “Net Zero” is important – but it is not the only factor, all (all) the policies of the establishment in the United Kingdom are economically and socially (culturally) damaging.
One can debate the motivation – are they deliberately trying to destroy the country they claim to “love”, or is the fact that the economic and social policies the establishment support are destructive an unfortunate consequence that they, the establishment, do not understand?
But whatever their motivation, whether they know it or not, the economic and social (cultural) policies of the establishment are destroying this country.
It is unlikely that the United Kingdom will make it to the next General Election in 2029 – the situation is that serious.
No Fraser Orr – the policies being followed (whether on Net Zero or anything else) are nothing to do with politicians advancing their “wealth and power” – if only they were.
Whether it is “Net Zero” or “Diversity” or run-a-away government spending – it is always from principle, insane-twisted principle, but still principle. For example, when Prime Minister John Major said “we have spent more money than Labour Party even promised to spend!” in his twisted mind this was an achievement, something he was morally proud of (his moral principles being inverted – horribly twisted). No one voted Conservative for higher government spending, and it did not benefit Mr Major in any way – but he thought it was a noble thing to do.
If it was the desire for bribes and positions of authority it would be vastly better – vastly less harmful, as bribes and positions could be offered to stop these policies. Sadly this will not work – as the people who rubber stamp the insane policies think they are doing good – they would be horrified and indignant if offered bribes to stop destroying the country.
And it is often the case that the politicians do not even know what the policies they rubber stamp really are – for example how many Members of Parliament understood the contents of the Environment Act or the Equality Act – both passed in 2010?
Very few. Most Members of Parliament just thought as follows….
“Social Reform” (bigger and more interventionist government) “is morally good, these Acts are Social Reform, so these Acts are good”.
@Fraser
Hexadecimal is a breeze, binary is a bit (get it) more of a faff.
@David Roberts
Hexadecimal is a breeze, binary is a bit (get it) more of a faff.
LOL, reminds me of programmer joke number 62,719:
A data packet walks into a bar.
Barkeep says “Gosh you look terrible!”.
Packet replies, “yeah, I have a parity error.”
“Oh”, says the barkeep, “I thought you looked a bit off.”
I can recommend Alex Epstein on the whole climate debate. He’s as good as Bjorn Lomborg.
Fraser,
The point I was making was that any industry that relies heavily on the price of electricity will move. You want a datacenter in the Western USA then the fact the sparks are about 1/3 the price in NV compared to CA is a factor. In terms of the UK, electricity is too pricey for us to become the “renwables powerhouse” because the manufacturing costs are too steep. Ed Milliband is hoist on his own petard. In the UK we simply can’t compete with China (and the rest) in terms of making this clean stuff because by going net-zero we are pricing ourselves out of the market for even making green stuff which, of course, requires cheap electricity. It is cutting your own legs off to climb a ladder.
Space… Yeah, right. OK, SpaceX has dramatically reduced the costs of launch but they are still enormous. Don’t get me wrong, I’d love to continue this discussion with you with umbrella drinks by the methane seas of Titan but… Until we get stuff like Skylon* and space-els we’re gonna have to put that date on hold.
Though, God help me Fraser, I’d just love it!
*Why the fuck did the UKGov stet Skylon and then spend billions on HS2? Yes, I know that the gov shouldn’t be allowed near anything more techie than a digging stick, but…
Hex is easy. Bin is too long-winded but with Hex you can specify an RGB colour in six characters. It’s just 000000 and FFFFFF.
The UK in particular, the West in general, and that includes my own poor; country seems to fall under the DSM-5 ICD-10-CM Z91.5. It is questionable if it is treatable, let alone curable.
John Enoch Powell in the 1960s, correctly warned of the consequences of mass immigration and the natural increase of these populations, he was the most popular politician in Britain – the public (not yet indoctrinated to hate themselves and hate their nation) overwhelmingly supported the warnings of Mr Powell – so, if Fraser Orr is correct about the motivation of politicians, people in politics would have rushed to copy Mr Powell – in order to gain popularity, and thus “money and power” – but they did the opposite, they denounced Mr Powell.
When the Conservative Party came to into office in 1970 it did things that no one had voted for – such as decimalise the currency (getting rid of coinage that went back many centuries), destroy traditional counties (some of which went back to Anglo Saxon times) creating a total mess in local government (just as the Labour Party would have done – and the Liberals, because “local government reform” was POLICY – see later), rush to join the European Economic Community (as the European Union then was) – i.e. GIVE AWAY POWER.
People such as Prime Minister Heath, and Prime Minister Wilson, and Liberal Leader Mr Thorpe, were utterly uninterested in “money and power” – they just wanted to be “Progressive” to slavishly follow ideological fashions, they did what they were told to do.
It is not different with “Net Zero” – it is nothing to do with politicians wanting “money and power”, it is everything to do with politicians rubber stamping the decisions of officials and “experts” – because the politicians have little real power and do as they are told.
At the local level, various nice buildings in Kettering were got rid of, starting in the 1960s, and replaced with horrible buildings.
I knew some of the local councilors who were “in charge” back then (they were still alive a few years ago) – they agreed that the buildings they destroyed were good and the buildings they replaced them with were bad, so why did they do it? It was nothing to do with “money and power” Fraser Orr – they “had to do it” because “it was policy” – who makes “policy”? No-one-knows. Councilors, like Members of Parliament, formally vote for “policy” – but they do NOT create it.
I later became a councilor myself – first a Borough Councilor and later a “Unitary Authority” Councilor, and it was just-the-same-in-my-day.
I have often talked to Members of Parliament – and Parliament is just the same, no one knows where “Policy” comes from.
“Net Zero” is POLICY.
For Mr Powell “it is policy” was never enough to vote for a measure – whether it was baby killing, mass immigration, the EEC, or anything else.
His resistance to “policy”, his demand that a measure be rationally defended by evidence and logical argument (rather than “this is what we must do” why? “because it is POLICY – it is the Progressive, Modern, thing to do”), was the real reason he was hated.
One of the most depressing things is that this slavishness is no longer confined to the political “elite” – many ordinary people now follow the fads and fashions, due to many decades of indoctrination for the education system and so on.
The Green Party would destroy this country very quickly – and they are the most popular party among the young.
“We have just four months. Four months to secure the future of our planet. The world is looking to Korea for leadership. I’m glad that the Chairman of the forum and many other speakers have used my campaign slogan “Seal the Deal” in Copenhagen. I won’t charge them loyalty
– Ban Ki Moon
Chief UN Global Warming mega $cam gravy train conformist asshat and dumbass simpleton, August 11 2009.
Or:
“The data doesn’t matter. We’re not basing our recommendations on the data. We’re basing them on the climate models.”
– Prof. Chris Folland,
Hadley Centre for Climate Prediction and Research
“Isn’t the only hope for the planet that the industrialized civilizations collapse? Isn’t it our responsibility to bring that about?”
– Maurice Strong
Founder of the UN Environment Programme (UNEP)
“To capture the public imagination, we have to offer up some scary scenarios, make simplified dramatic statements and little mention of any doubts one might have. Each of us has to decide the right balance between being effective, and being honest.”
– Dr. Stephen Schneider,
Greenhouse Superstar / Leading greenhouse advocate, in an interview for “Discover” magazine, Oct 1989
“This is the first time in the history of mankind that we are setting ourselves the task of intentionally, within a defined period of time, to change the economic development model that has been reigning for at least 150 years, since the Industrial Revolution”
– Christiana Figueres
Executive secretary of U.N.’s Framework Convention on Climate Change.
Getting the drift, yet?
@NickM
Space… Yeah, right. OK, SpaceX has dramatically reduced the costs of launch but they are still enormous. Don’t get me wrong, I’d love to continue this discussion with you with umbrella drinks by the methane seas of Titan but… Until we get stuff like Skylon* and space-els we’re gonna have to put that date on hold.
Titan and Saturn are rather a lofty goals, but low earth orbiting satellites? Obviously not difficult for SpaceX. The planned launch cost for Starship* is about $10 millon per launch with 100tons to LEO. I don’t know how much a data center weighs but surely you can get a pretty substantial one in 100 tons and add a little booster to get it to Sun sync orbit and you are golden. SpaceX has spent 20 years preparing for this and Google thinks they can buy a rocket company and catch up? No, they have the delivery vehicles, massive experience (they have more satellites in orbit than the rest of the world put together), they have massive experience building data centers — which they do at ridiculous speeds — and already have, in Starlink, and existing network infrastructure.
For a launch cost of $10million they are buying real estate and free power. Honestly, I think the launch cost will only be a small fraction of the cost. These data centers cost billions of dollars to build here on earth. Plus none of the regulatory hassle, or problems getting enough power. TBH I think in the next ten years there will be a gold rush for that space. There is surprisingly little space in sun sync orbits if your satellites are deploying multiple kilometer wide solar panels.
So for most companies orbital data centers are a far off pipe dream, but SpaceX/xAI are absolutely primed to do it. They have vastly more experience at it than the rest of the world combined, they have the hardware, the expertise, the people and the money to make it happen. This is surely why SpaceX bought xAI and why SpaceX is going IPO.
So it is not some SciFi novel fantasy. I think we will see most compute in the sky five to ten years from now and I think most of it will be owned by some combination of Musk Inc.
BTW, there are concerns about maintenance, I’m pretty sure they will leverage their expertise in robotics to do this. But no reason there couldn’t be a human compatible module on there and people could actually go and work in one of these data centers. Again, SpaceX has extensive experience on this. It would be the first time people went to work in space not just to screw around with fake experiments. I think this is unlikely, but it is cool to think about.
It was one of Sir Terry’s characters who said that the best kind of war was one where the enemy doesn’t even realize they are being attacked. This seems to me to be exactly the situation in the west where hardly anyone grasps that Western Civilization is being deliberately destroyed by its enemies.
Although talking about Europe in general, this surely also applies to Britain. This stark diagram compares the number of > $10billion companies formed in Europe in the past 50 years.
https://x.com/FutureJurvetson/status/1864066992917102874
And that’s compared to the USA which is becoming more and more sclerotic and bureaucratic by the day.
Fraser,
I hope you are right but… “a little booster to get it to Sun sync orbit” is not the small issue you seem to think. Short version is that data centres on this planet are much easier. For off-world there is of course the issue of radiation-hardening them outside of the megnetosphere amongst other things. The only advantage to sticking them in space is to avoid governmental (with the accent on “mental”) interference (bad) or help (worse) which is admittedly a big deal. A sane solution would be to build here along with SMRs and the “waste” heat used to heat houses pro-bono. That is not just sensible (though it is) but also saleable (which matters a lot as far as opinion goes) especially if you get the media to show granny luxuriating in January and compare and contrast with her frozen mummified corpse in front of the one bar electric fire that failed due to the windmills. Shorter version: unless we get people to fall in love with fission we is fucked.
Bruce and Rour le Jour – yes.
It was clear even at the Rio Conference of 1992 that the Carbon-Dioxide-is-evil theory was just an excuse for what the international establishment wished to do anyway – world “governance” on a sort of Henri Saint-Simon Corporate State form of Collectivism.
If “Policy” has a human source (rather than “Policy” just being vomited up from Hell – or whatever) – that source is the international establishment.
“Diversity” (“Diversity” is a very different thing from ordinary diversity) has the same purpose as the Carbon-Dioxide-is-evil doctrine – namely the death of the West, the extermination of what is left of liberty.
To answer the question of Fraser Orr – some politicians do not know that “Policy” is designed to destroy the West – I am sure that Prime Minister “Boris” Johnson had no idea (none – none at all) that the Policy he followed, on mass immigration, on “Net Zero”, on wild government spending, and so on, was designed to destroy this nation.
But some politicians DO know – for example the leader of the British “Green” Party knows very well that these doctrines (“Diversity”, “Net Zero”, wild government spending and taxation, and so on) are designed to destroy the West – he knows and he supports that objective, he wishes Western civilization (which he hates) to be destroyed.
And the Green Party is the party that is most popular among the indoctrinated young – and not so young.
@NickM
I hope you are right but… “a little booster to get it to Sun sync orbit” is not the small issue you seem to think. Short version is that data centres on this planet are much easier.
Are they? I live in a city that has a gigantic tech corridor full of the top labs for networking the phone companies. It has packed with technology companies. The city is pretty conservative for Illinois, and is very pro business. Recently someone tried to open a small data center here and it got rejected because of a mass campaign or luddite ignoramuses. Even in Memphis xAI is dealing with constant pushback from the people as they try to expand their offering (including the aforementioned need to build their power plant in a separate state.)
There are some challenges to building in space, but the difference is that these challenges are engineering problems entirely within the control of the builders, not social problems subject to the whim of the voters and leftie anti tech movements. And if you think your idea of nuclear reactors to power these data centers will be less controversial and difficult to get through the anti-everything mob, I’ll have to assume you are not NickM but an AI posting as NickM, suffering from a hallucination 😉.
Regarding radiation, certainly it is a bit of a problem, but less than you might think. The orbits would be below the Van Allen belts, and one unusual thing about AI data centers is that they are pretty robust to random bit flips. I think that there is a bit if a risk of ablation of the solar panels from micrometeorites, but that is more in the nature of a consumable. And regarding orbital height — I think this is why early adopters will get all the premium orbits below the van Allen belts.
In the longer term I think even these are usable when we can launch the satellites from the moon, and use lunar material for bulk radiation shielding, with of course the advantage that it is much easier to launch mass from the moon than the earth.
Which all sounds like crazy science fiction. But the reality is that orbital data centers are certainly a 5-10 year project, and the moon thing? a 30-50 year project. So maybe I won’t be around for it, but my kids will.
There is no need for the project that Fraser Orr describes.
@Paul Marks.
There is no need for the project that Fraser Orr describes.
How short sighted of you.
Let me explain why you are wrong in one word: Macrohard. Most people have not heard of this project. It is ongoing but I think it is probably one of the most significant things happening in the world today.
What is it? Macrohard (a pun on Microsoft) is an attempt to create a software company using no people, only AI agents. He picks on Microsoft because their only product is digital. The idea is that, since the input (customer desire) and the output (software products, marketing and a store) are all digital, it can be done entirely by AI agents. Marketing, product design, product management, software design and creation, QA, product evaluations, accounting, legal, compliance. All done by AI agents. Which is to say you can reproduce the whole of a 3 trillion dollar company in a data center with no people, and no costs beyond data center costs. And if you can do it once, you can do it many times.
FWIW, I have seen some of the early products, video games produced entirely by AI that are apparently excellent (I’m not a gamer so would not know. I’ll ask my kids.) High end video games typically take 12 months and huge teams of engineers to produce. AI produced this over the weekend with no people, just AI and electricity.
Now, I’m not talking about replacing Microsoft itself really, but a company that creates new and useful products never thought of before, just as Microsoft did.
But surely this is just a niche idea? Not at all. Five of the ten largest companies in the world produce only digital products. Microsoft, NVIDIA, Apple, Google and Meta*. When this project succeeds it means that AI is able to generate companies of that scale, and do it quickly, repeatably all in software with tiny overhead. This means that it is possible to translate a few million dollars in data center costs into hundreds of billions of dollars in value.
What about the other huge companies? What about Amazon? Consider what Amazon does: really Amazon is two businesses an online retailer and an IaaS service as AWS. As an online retailer they do basically two things: marketing to connect buyers and sellers, and logistics to deliver the product from the buyer to the seller. The whole marketing side can be done entirely by AI because it is all software. What about logistics? In this case it can be done entirely by robotics and self driving vehicles. Imagine a warehouse where receiving, picking and packing is all done by Optimus robots. And imagine delivery center to delivery center transfers by self driving Tesla semis with Optimus robots to move the product. And finally imagine last mile delivery done by Tesla Cyber trucks with an Optimus robot to take the product from the truck to the door.
As to IaaS, this too is composed of two parts: a large software stack, which AI could certainly produce, and actual physical hardware, which of course, with massive data centers, could be provided. One thing here is that this is probably not suitable for orbital data centers, latency probably demands that these IaaS data centers be terrestrial (I could say a lot more about this, but I’m sure it is getting boring.)
So what does all this mean? It means that AI can produce massive value in the economy, and when enhanced with some capital for robotics, self driving vehicles and warehouses, it can produce massive amounts of the existing economy.
And why does this matter? Because the ability to transform a few million dollars of investment into multibillion dollar companies in a few short months means that our nation’s economies can grow at spectacular rates. It means that the thing that keeps you up at night, the national debt, can be wiped out by one year of 600% GDP growth. Not in Britain of course because of Net Zero, but other places.
What does that world look like? I have no idea, it’ll be a massive transformation, but for sure it’ll make the world vastly richer, bristling with new value, new products, new ideas, new services beyond our imagination. How that wealth is shared among the population will surely be one of the great political debates on the 21st century.
BTW you might not like this. And for sure I agree there are some downsides. The turmoil in the economy will be the biggest since the industrial revolution, maybe bigger. But it doesn’t matter whether you or I like it, it is coming, like, to quote Musk “a hypersonic tsunami”. And certainly better that Elon Musk is in charge than Xi Jinping.
[*] BTW if you are saying: “Fraser, Apple makes phones, NVIDIA makes chips”, but actually they don’t. Foxconn makes phones, TSMC makes chips. Apple is a company that produces phone designs, NVIDIA is fabless, they produce chip designs.
“I think there is a world market for maybe five computers.”
– Thomas Watson, president of IBM, 1943
“Television won’t be able to hold on to any market it captures after the first six months. People will soon get tired of staring at a plywood box every night.”
– Darryl Zanuck, executive at 20th Century Fox, 1946
“Nuclear-powered vacuum cleaners will probably be a reality within ten years.”
– Alex Lewyt, president of Lewyt vacuum company, 1955
“There is no reason anyone would want a computer in their home.”
– Ken Olsen, founder of Digital Equipment Corporation, 1977
“Almost all of the many predictions now being made about 1996 hinge on the Internet’s continuing exponential growth. But I predict the Internet will soon go spectacularly supernova and in 1996 catastrophically collapse.”
– Robert Metcalfe, founder of 3Com, 1995
“There is no need for the project that Fraser Orr describes.”
– Paul Marks, 2026
And what games would those be? Are you referring to procedural generation of environments like “No Man’s Sky” or something else?
@Perry de Havilland (Prague)
And what games would those be? Are you referring to procedural generation of environments like “No Man’s Sky” or something else?
Sorry, I don’t know. I’m not really a gamer person, I’ve just seen videos and heard other people who know about these things talk. But I’ll try to dig up what I saw. To my eyes it was pretty spectacular, and not just an “environment” but a complete game from the ground up.
Fraser Orr – there are many sources of power, there is no need for the project you describe.
In case anyone doubts the evil of the “international community” (the international Corporate State) I refer you to an article in this week’s issue of one of their pet publications – the Economist magazine.
The Economist magazine declares that the Covid “vaccines” saved vast numbers of lives – this is part of an article smearing RFJ Jr – and also to trap President Trump (supposedly this toxic muck was the “greatest achievement” of his first term – thus trying to trap him into defending the toxic muck).
The Economist magazine (whoever this year’s version of “Lexington” may be – the Economist speaks with one voice, all articles in it are checked before publication) knows perfectly well that these injections were not “vaccines” and did NOT save vast numbers of lives – indeed these injections injured and killed people.
They do not care about human life – they care only for the international Corporate State.
Let that sink in.
Procedural generation…
Elite was done like that in the ’80s… The various follow-ups also used it. It makes them compact pieces of software but tedious as Hell. This is not new.
Fraser,
I don’t believe space is easier to conquer than NIMBYS. The latter are slowly coming round to nuclear. Note my previous quote from the President of France. Just two other things before I go… Space is cold but it ain’t great for cooling because there is nothing there. Computers need cooling. Secondly, unless we have plentiful cheap electricity here on Earth your Utopia is going to remain a no-place.
Not “new” in the sense an F35 is not something “new” compared to the Wright Flyer I suppose, they are both aircraft.
@NickM
I don’t believe space is easier to conquer than NIMBYS.
I’m sure we will find out. Let’s see where SpaceX/Tesla stock price is compared to Meta five years from now, and see which does better. I’ll meet you back here. Stakes are one beer. And not one of those cheap watery lagers, a good pint of dark ale.
The latter are slowly coming round to nuclear. Note my previous quote from the President of France.
They haven’t tried to actually build one yet. So I’ll hold judgement on that for sure. FWIW, France has always been pretty pro nuclear power. A large fraction of their generating capacity is already nuclear, I think maybe 70%?
Space is cold but it ain’t great for cooling because there is nothing there. Computers need cooling.
I’ve heard this argument before, and I don’t think it holds up. These data centers will have very large solar arrays and the back is always in shadow, so it would be easy enough (“easy” is obviously a relative term 😉) to put radiators back there. As a data point, the radiators on the ISS are one third the area of its solar panels (and that is a suboptimal cooling environment since they are not always in shadow.)
Secondly, unless we have plentiful cheap electricity here on Earth your Utopia is going to remain a no-place.
Not sure I ever claimed anything would be a Utopia, just quite different.
FWIW, I hate to go on about Tesla (actually that’s not true, I’m a shareholder so trying to talk it up to boost my retirement fund 😉) but their megapacks go a long way to helping. Peak capacity is double average capacity because of variable load, these huge batteries buffer the generating capacity effectively doubling the capacity of the power generation stations. And they do it without destroying beautiful valleys in the mountains. I think over the next few years we will see local substations converted to megapacks — they also handle the down voltage conversion — meaning higher quality, more reliable, more distributed power, less at risk from natural disasters. This latter point is important, both because power capacity is distributed and because in case of failure they can truck in a new fully charge megapack to keep the lights on even if the HT power is down. And power stations can run at optimal continuous load, not be constantly turning on and off.
I mean even the crazy people in Westminster can surely get behind that?
@Perry de Havilland (Prague)
Not “new” in the sense an F35 is not something “new” compared to the Wright Flyer I suppose, they are both aircraft.
Made me LOL. Sorry though, I don’t know much about the gaming world. I still think my Wii is pretty cool.
As long as we don’t end up with an AI that promises ‘to serve and obey and guard men from harm’…
Is this the future?
In order to live, most families in the developed world require a monetary income. The sources of this income are typically: a job, acquired income (mostly pensions), or provision from the state. The utilization of AI is predicted to replace a high proportion—perhaps even most—of these jobs.
This displacement of labour has been occurring since at least the time of the Tolpuddle Martyrs and through the automation of the twentieth century. In the main, and despite some painful adjustments, human ingenuity has always generated new, different replacement jobs for the majority of people.
This may yet be the case with AI. However, it is also possible that most families will become dependent on the state for their income. Leaving aside the psychological impact this would have on individuals and society, where will the state get the money to provide an income for such a high proportion of its population?
The first answer that comes to mind is the taxation of the organizations whose AIs provide the goods and services society requires to survive. Yet, this appears to result in a society of diminishing returns. The justification for replacing people with AI is predicated on the vastly reduced cost of the goods and services required for people to live.
The situation for the organizations providing these goods and services becomes farcical. They finance the creation and maintenance of AI production systems, expecting—over time—to recover these costs plus a profit.
Who is going to buy the goods and services thus produced? The general population, of course. Where is the general population going to get the money to buy them? The state, of course. Where is the state going to get the money to give to the general population? The AI organizations, of course. And where are the AI organizations going to get the money to pay their tax bills?
This merry-go-round will initially be supported by the diminishing number of people with jobs—diminishing because AI will constantly improve its efficiency and expand the roles it can fill. The system cycles down. Consumers cannot afford the prices of their needs, so prices must drop. Consequently, the AI companies cannot afford to pay their tax, and the state cannot afford to provide an income for its citizens.
Round and round we go, until the prices of goods and services reach zero.
The AI companies are now working only for love.
Presently nuclear fission is the best non hydrocarbon source of energy (although there is nothing much wrong with hydrocarbons – coal, gas and oil, which Britain has, insanely, turned against – and President Trump, wisely, has not), most likely small “modular” nuclear reactors that, for example, Japan is going for.
However, in the long term nuclear fusion will be the main power source – and, before Fraser Orr jumps in, I do know know that the sun is a form of nuclear fusion – but it would be better to generate nuclear fusion power here.
@David Roberts: I’ve been wondering along the same lines.
Let us take a company producing a good or service. It employs 1000 people to do this. It brings in AI and manages to produce exactly the same amount of goods or services with half the employees. Its profits rise by the difference between the cost of the sacked 500 employees and the cost of paying for the AI that replaces them. Even if the State then taxes the company 100% on its extra AI generated profits it doesn’t get enough money back to pay all the sacked employees a UBI anywhere near what they used to earn. Apart from which its highly likely that the market price of the good or service will fall, as its now cheaper to produce and thus competition will drive its price down, and the company won’t make all the extra profit it thought it would. And if the State then expropriates all the AI related profits why bother using it in the first place? There doesn’t appear to be any extra profit created by the system to tax to pay everyone to do nothing. Especially if a lot of the AI revenue is disappearing away from any given country to another.
All AI seems to promise is make doing existing things more cheaply (mainly by reducing human labour). Previous leaps in technology have solved problems that seemed insoluble – electricity made TV and radio communication possible, the internal combustion engine allowed us to fly, the microchip allowed us to create devices that can do things that would seem like black magic just a few decades ago. AI by comparison just promises to replace Judy in accounts with a data centre based in Nuneaton. Previous new technologies opened up entirely new vistas of human economic activity. AI seems to offer none of that, and just promises to change the size of the slices of the pie without making the pie any bigger overall.
Fraser,
You mention the ISS. Now isn’t that a case of $150 billion for the worst caravan holiday ever? Until launch costs per kilo are reduced by several orders of magnitude space is just for extremely high value stuff such as comms, recon (inc. weather), science and Katy Perry. Basically only things that need to be in space. I have never heard a realistic business model put forward for anything you can build on the deck. A fully re-used SpaceX Starship is still about $200 per kilo to even LEO. That gets down by two or three orders and then I’ll stand you a pint of whatever your heart desires. For me it’ll probably be an IPA. I shall look forward to it because I can’t lose. Either you buy me a drink or my wildest dreams come true and I seriously doubt your tipple of choice will, anywhere on this planet, cost more than it would up there! So, either way we shall have a pint. It may be in Cheshire, or if I lose, at Lagrange 5! Either way I shall look forward to it.
I hope to lose.
@Jim and @David Roberts, I think it is hard to predict what will happen, not so much with what the technology is capable of, but because it is rather dependent on how people react to the inevitable changes. But here is what I think:
Let’s start with the assumption that I am right and AI based companies really can transform a few million dollars of compute time into billions of dollars of value: remembering that projects to do this are already underway, and are perhaps only four or five months in. Now, @Jim presents a rather sad and limited approach to this of existing companies trying to adapt, but I don’t really think that is where the action is. Honestly, I think these traditional companies will fade away or hyper niche. Why? In a sense for the reason Jim gives — they just can’t possibly keep up with companies that have no payroll, and workers that work a hundred times faster than people, 24/7, and don’t even take pee breaks.
Now let’s be clear, these AI companies are not just replacing existing industries, though they are probably doing that, they are creating whole new industries out of whole cloth just as Apple or Microsoft or Facebook did. Just to give a taste of this — in the maybe two years that AI has been working in drug discovery it has discovered at least 20 new drugs including two new classes of broad spectrum antibiotics that are very effective against MRSA and gonorrhea. So we are talking about new industries that make life very much better, created out of thin air — or at least created out of silicon and electricity.
So if companies are translating millions of dollars into billions of dollars in a few short months, what does that mean? It means massive levels of currency deflation. If a million dollars can make a billion dollars of value, that is a huge deflationary force. Now, me personally, I’m not opposed to deflation, I think, like inflation, it has pros and cons. However, the government-industrial complex thinks deflation is an absolute unadulterated bad. And the almost universal approach to deflation is printing more money — in this case a LOT more money.
Now let’s combine this with the political reality of mass unemployment, people on the street with torches and pitchforks. I think there is only really two political possibilities: either the government takes some of that freshly printed money and gives it to the mob — which is to say introduces a universal basic income at a high level — or alternatively they try to put the genie back in the bottle by banning AI companies.
The problem with the latter approach is that if they try that the AI companies can simply move to a friendlier location, like Dubai or Singapore, both of which would happily accept them. And if those data centers are in orbit, this is made considerably easier. So banning AI simply makes everyone in your country dramatically poorer and dependent on small city states that will quickly become vastly wealthy. FWIW, I think it is a miracle that Musk is still in the USA. If it were me I’d have moved away from high tax, high regulation America a long time ago.
So my conclusion? AI takes over most productive work and makes for a vastly richer world full of goods and services we never even imagined. And the government extracts some of that profit through either money printing or taxation, and provides people with a UBI at a fairly high level. This means that people have to switch their thinking to find other ways to find purpose and meaning in their lives than their job.
BTW, I’m not by any means arguing that this is a good thing, I’ve always been opposed to a UBI, what I am saying is that the trends in society, economics and technology seem to me to lead inexorably to this conclusion. And I am also not saying it will happen next year, or even this decade. It will be a tumultuous transformation and lots of things can go wrong to make it worse.
But it is what I think the future holds — a world where people try to move up Maslow’s hierarchy of needs, where jobs are a lot less important, where money is a lot less important, where people try to find meaning, purpose and advantage in other areas of life. And, I think unfortunately, a world where, as wards of the state, there is a lot more government interference in people’s daily lives.
@NickM
You mention the ISS. Now isn’t that a case of $150 billion for the worst caravan holiday ever? Until launch costs per kilo are reduced by several orders of magnitude space
But Starship DOES reduce launch costs several orders of magnitude. The current iteration is planned to launch 100 tons to LEO for a $10 million launch cost. That’s $100 a kilo. With Raptor 4 and more experience the plan to switch that to 200 tons at a $2 million launch cost which is $10 a kilo. The Space Shuttle (even in unadjusted for inflation pricing) launched about 30 tons for $500 million, or $16,000 per kilo, which is quite a few orders of magnitude improvement.
Hopefully “AI” is at least partly a scam (a confidence trick)- but if it really is possible to produce an artificial “intelligence”, i.e. a free-will being, then we are trouble – “Skynet” (“Terminator”) style trouble.
But be that as it may – there is still no need for some power station in space (other than to power space stations and colonies in space) – delivering energy back to Earth, it is far more economical to produce electrical power on Earth – either via coal, gas and oil, or via nuclear fission (perhaps small modular nuclear reactors), and, in the long term, nuclear fusion – which would have no nuclear waste for anyone to complain about.
As for the economy – obviously, given run-a-way government spending and the Credit Money monetary and financial system (Credit Bubble banks and all), the economy will decline, if not collapse, over time – at least in countries such as Britain – where the vast majority of people either work for government (in one form or another) or depend on benefits (a majority of people get benefits – if we include state pensions, which are benefits – regardsless of how angry it makes people to say that, there are no investments funding it – none, and then there also the many millions of people who work for the state, in one way or another).
The idea of some massive economic improvement is, sadly, nonsense.
There will be decline – or collapse, in most Western nations.
Also, at least in Western European nations, it is sadly, tragically, inevitable that, at some point, the ethnic groups will fight each other – as they eventually did in Lebanon (due to demographic change) – this conflict will also have bad consequences for the economy.
What will happen in the United States – I do not know.
It takes 50 kg of fossil fuel to lift one kg of payload to space. No technological innovation is gonna change this fact.
AI is, now what?
What is to be done in a world where AI results in a larger and larger section of society living without paid employment, requiring some form of Universal Basic Income? My answer is: education, education, and education.
My answer is based on three presuppositions:
1. That the environment for the creation of AI must be actively organized to promote a vast variety and multiplicity of systems.
2. That in an open and democratic society, the better each citizen is educated, the more that society will progress.
3. That all people who receive an excellent education benefit from it, thereby enhancing both their own lives and their contribution to society.
Given these three presuppositions, the new AI technology must be utilized to provide a bespoke education service at minimal to no cost. Tailored to each individual, this service should be as enjoyable—and perhaps even as addictive—as the best modern media. It should be driven across the internet by many diverse and adaptable AI tutors.
What else is there to do? Nothing. I expect a highly educated populace to be more than capable of meeting the challenges of a new AI-driven world.
You can roughly divide the world of work-producing-value into the people working with tangibles – farmers, mechanics, hitmen, plumbers, house builders, factory assemblers . . . and people working with intangibles – lawyers, teachers, software people, writers, accountants . . .
A.I. is going to wipe out the livings of the second group. Those in the first group are going to be doing better. We still NEED what they produce.
So now we’re going to watch as the second group – the intangibles – figure out how to keep calling themselves the cream while living off of the production of the tangible people.
It will be like the original establishment of the commodities central markets all over again, but will involve . . . everything produced.
There’s going to be a lot of re-thinking about where a smart and promising young person should aim their career. I doubt the answer will be “college.”
A.I. is going to wipe out the livings of the second group.
No, it is merely going to create a different kind of the second group (the intangibles). That is in no way meant to denigrate the value – present and future – of the first group, although their work is likely to change as well.
Yes Jacob – and even if there are colonies on the Moon, and so on, in the future – they will NOT be sending electrical energy back to Earth, they will be using the electricity they generate for their own needs.
As for “AI” – if something is a true “intelligence”, a free will being, then it is a someONE not just a someTHING – and we have no right to order this Artificial Intelligence to do anything.
Old Jewish stories of Golems spring to mind – but in this case they would not be like humans in body, but rather in MIND.
And you know that how?
Paul:
Very true. And not just the right, but probably not real ability either.
However, I don’t think anyone here is using the term AI in that sense, at least not in the current context. And in any case, intelligence and free will are two very different things.
@bobby b
You can roughly divide the world of work-producing-value into the people working with tangibles – farmers, mechanics, hitmen, plumbers, house builders, factory assemblers . . . and people working with intangibles – lawyers, teachers, software people, writers, accountants . . .
Sorry BobbyB you are only looking at half the equation. It isn’t just AI, it is AI and robotics. Most of these other jobs can be done by robots, though I’m pretty sure some sort of “don’t kill a human” rule will be built in there, so you’ll have to do your own hitman work.
@Paul Marks.
they will NOT be sending electrical energy back to Earth, they will be using the electricity they generate for their own needs.
You keep banging on about this. I have never suggesting that orbital power stations will send electricity back to earth, it’ll be use in orbit. I don’t think sending energy from space to earth is even particularly practical or desirable.
And I agree with Alisa’s comments on your other points. AI as an “artificial general intelligence” whatever the heck that means, is really a distraction from its actual disruptive power.
David Levi – since the thread is about AI, I asked Grok: the potential energy of a kilogram in LEO is much smaller than its kinetic energy (speed to retain orbit), so we’ll ignore the potential energy part. That kinetic energy is ~30 MJ/kg. The chemical energy in 50 kg of fossil fuel is (again, according to Grok) in the range of 24-43 MJ, depending on type of fossil fuel. Of course it takes a LOT more energy than that to lift a kg of anything to that LEO, due to engineering inefficiencies that SpaceX is trying to minimize.
Yeah, probably. I’m just not seeing it yet. I can see a robot on a line. I cannot (yet) see a robot walking about and fixing or building things in non-controlled environments. Luddites R’ Us, I guess.
But I can already see the marketing schemes fighting this trend. “Built by humans, for humans . . .” Non-robotic build will assume the same level of trendy aura as “organic” does now for food.
So, in this UBI-but-not world, who will decide who can afford the new luxury auto, and who has to settle for the Yugo? If we’re not “making money”, what differentiates us as having more value – or simply more to spend – than that layabout over there?
Tribe? Friends? Political pull?
I always liked the system of “reward for providing value” resulting from a market economy. I am not sure I will love the system where my value versus your value is determined by some other factor or decider.
@bobby b
Yeah, probably. I’m just not seeing it yet. I can see a robot on a line. I cannot (yet) see a robot walking about and fixing or building things in non-controlled environments.
Apologies, I was not clear. I don’t mean robots now, I mean specifically Tesla Optimus. The robots today are very limited. And, to be clear, the first few versions of Optimus will also be very limited, but once they have five years under their belt they will be transformative. And to be clear, no other robotic company can come close to what Tesla will do, because a huge amount of what is needed to train Optimus is cross over with Tesla full self drive. And because success is based on how much training data you can acquire, once you are in the lead and assuming you know what you are doing, your lead will grow exponentially.
Soon enough almost everyone will have a robot in their house and it will be perfectly capable of doing all the electrical, plumbing, building work etc. for you. Want a new bathroom? Tell Optimus and buy the raw materials. Want a pool in the backyard. Optimus will do it. Flip this dilapidated house for profit? Borrow your buddy’s robot to work with yours, and they will work 24/7 till it is done. I have been thinking about painting a room in my house for about three months and haven’t gotten around to it. If I had this robot I could tell him to do it one evening and it’d be done when I woke up the next morning.
There is a place of friction though, namely where human laws prevent robots from doing stuff. For example, unionized labor will be very resistant and that will hold up a lot of adoption. Also, some professions, doctors, lawyers etc. will also be held up, not because they can’t do a better job, they can or they will be able to do so. It is just the human laws require an actual human with appropriate credentials to do these jobs. I wonder if I can represent myself in court pro se, and bring along my robot to help?
“Built by humans, for humans . . .”
You could well be right, and it might be good business to be in, however niche. And I think there will be a lot of those sorts of “resistance”. It is why I said earlier that the technological progress is fairly predictable, what isn’t clear is how humans will react to it, resist it or embrace it. But, to be clear, it is coming, and it is coming fast.
So, in this UBI-but-not world, who will decide who can afford the new luxury auto, and who has to settle for the Yugo? If we’re not “making money”, what differentiates us as having more value – or simply more to spend – than that layabout over there?
Don’t know. We will have to wait and see. But what I’d suggest is that money and jobs will be a lot less important, and we will have to find new ways to evaluate ourselves, differentiation ourselves and advance ourselves than these mechanisms. But humans are pretty good at that kind of thing. Watch a couple of kids, one proving he is better than the other because he finished his apple sauce first, or because he can jump higher, or has read more books, or whatever. I assure you, we will find a way to compete and “win” even if money and jobs aren’t a measuring stick anymore.
I always liked the system of “reward for providing value” resulting from a market economy. I am not sure I will love the system where my value versus your value is determined by some other factor or decider.
I do too, and I’m not advocating for this system. I’m just saying that is what I think is going to happen, and it is a future with both pros and cons. As to your question about cars: I think the answer is that both people get the luxury sports car, and then have to find some other way to measure their penis 😉.
I don’t know about the other authors in this parish, but I intend to promote my writings as being the result of fully human creativity, with all its foibles and limitations. Dunno if that will be a marketing advantage or disadvantage, but I’m writing to please myself, and would feel that I was slighting myself to rely on AI at any level beyond a grammer (sic) checker.
“Now let’s be clear, these AI companies are not just replacing existing industries, though they are probably doing that, they are creating whole new industries out of whole cloth just as Apple or Microsoft or Facebook did. Just to give a taste of this — in the maybe two years that AI has been working in drug discovery it has discovered at least 20 new drugs including two new classes of broad spectrum antibiotics that are very effective against MRSA and gonorrhea. So we are talking about new industries that make life very much better, created out of thin air — or at least created out of silicon and electricity.”
So who exactly is going to be buying the wonderful new products/services these ‘AI companies’ are producing? The millions of people who now live on a basic UBI because thats all they have? AI companies can only sell to actual people, and if all the actual people are unemployed there is no market for their products and services is there?
The AI boosters are trying to have it both ways – AI will at the same time make stupendous profits that will pay be taxed to pay for UBIs for everyone and they will also make products dirt cheap. Both can’t be true.
This is a debate that has been going on since at least the war and is satirized by sir Terry as “What if golems did all the work?”
Humans are naturally competitive. Every man wants a bigger house, a flashier car and a prettier wife than his neighbour. Society can either harness this competition, which is capitalism, which makes everyone richer, or try in vain to suppress it, which is communism, which makes everyone poorer. UBI is just communism wearing a different hat.
The question is moot in any case because outside the US the winner won’t be capitalism or communism, it will be Islam.
@Jim
So who exactly is going to be buying …? The millions of people who now live on a basic UBI because thats all they have?
I clearly stated a bunch of times that it will not be a “basic” UBI but a generous one.
AI companies create value. Some of the value goes to the owners, some of it is distributed to the people. As a simplistic example: AI produces 1000 bushels of wheat. 500 goes to the owners making them rich. 500 goes to the people allowing them to eat. That isn’t the way the transaction is conducted, but it could be (a sample) net result.
Please think about the difference between value and money. Money isn’t value, it is just a way of keeping score.
And, to once again repeat this: I am not advocating for this system, I am just saying that that is probably where we will end up, or something like that anyway.
Fraser Orr.
Even a basic “UBI” (i.e. giving people money in return for no-work) would be folly – it would lead to an increasing under-class over time, that is why Welfare Reform was essential – starting in Wisconsin and becoming more general – because it had to be followed nationally, welfare had to be reformed to get away from giving able bodied people money in return for no-work.
As for a “generous” welfare handout policy – that would lead to economic and cultural (societal) breakdown, fairly quickly.
“Money isn’t value, it is just a way of keeping score”.
That is NOT true Sir – if money is not a store-of-value, NOT just a medium-of-exchange, then it is no good. This is why money should be something that people choose to value before-and-apart-from its use as a money.
I am sorry Sir – but you are just wrong, not just on one thing, but on all the basic themes you have outlined in this thread.
As for “AI” – no it does not change any of the above.
Whether it is Star Trek or the “Dark Age of Technology” in the “Warhammer” background – some people think that fancy technology changes the laws of economics and the more general laws of human beings interacting in society.
It does NOT. As these laws of the interactions of human beings are universal – they do NOT depend on the “historical stage” any more than they depend upon race.
And this is true even if the fancy technology is real – rather than a confidence trick (a scam).
I think we can judge what the outcome of a massive UBI reliant society would be like by judging what we see from examples of welfare dependency and also others who don’t need to work due to having independent wealth they’ve inherited or have rich parents etc: there would be a tiny proportion who would flourish pursuing creative, physical and intellectual projects out of passion freed from the need to wage earn. Then there would be the mass who would vegetate and just indulge themselves in trivialities, vices and slop.
Martin – and it (money-for-no-work) gets worse over time, over generations.
UBI is an utterly demented idea (worthy of the totalitarians of the World Economic Forum and the United Nations) – and “AI” (even if real – rather than a-bit-of-a-scam) does NOT change this.
Nor is money “just a way of keeping score” – money should be a store-of-value not just a medium-of-exchange, money emerges when people use something that they ALREADY value, before-and-apart-from its use as money. For example, in the Middle East silver was used as money – thousands of years BEFORE the invention of coinage (people who confuse the invention of coinage and the invention of money – are just wrong).
Carl Menger and Ludwig Von Mises explained these matters a long time ago.
“AI companies create value. Some of the value goes to the owners, some of it is distributed to the people. As a simplistic example: AI produces 1000 bushels of wheat. 500 goes to the owners making them rich. 500 goes to the people allowing them to eat”
We produce wheat now, using humans (and machinery). How much EXTRA wheat is AI going to produce? I’d say none, give or take a margin of error, as the fundamental factors that drive wheat production (the weather mainly) are outside of its control. Hence my point, AI isn’t going to produce lots more stuff for the same (non human) inputs, its just going to make the same amount of stuff with the same (non human) inputs for less cost. That hasn’t created any extra value, because no more stuff has been made. AI can’t create any extra value, because behind every extra unit of production lies a unit of energy, and AI won’t make energy cheaper.
Lets ignore £££ and look at it in terms of goods. In your example above how does 500 bushels of wheat make the producers of it rich, if the masses are satisfied by their free 500 bushels? The owners retained share be worthless, because everyone who wants wheat already has it. So there won’t be any profit at all. Its the same for everything – if the AI car company makes so many cars so cheaply that it can give away enough cars such that everyone can have one for free then who do they sell the remainder to?
The AI Genie
The AI genie is out of the bottle, and it is not going away. The only remaining issue is: how will it be integrated into our society?
I see two primary paths. In the first, AI is used for the benefit of an elite group who manages to seize control of the technology, using it to subjugate the rest of us. This is the “1984” or “Terminator” scenario. In the second, its use is organized—somehow—for the benefit of us all.
Those who deny or ignore the existence of the AI genie only prevent themselves from influencing this integration.
My entries to this thread have all been edited by the free Gemini AI.
Perhaps we should do with AI what the Chinese do with green energy, flog it to other people but not use it ourselves.
@Jim
We produce wheat now, using humans (and machinery).
Sure but it was just meant as an illustration of the difference between value and money, and how it can be managed.
However, these AI companies may replace existing industries but will certainly create entirely new industries and products that we don’t have right now. I gave some examples above. And that is an expansion on the total “value” being produced. Since that value is produced with very small input, it is why, as I argued above, AI companies (by which I mean companies run by AI, not companies humans run to produce AI) will be massively deflationary, which leans to my conclusion about where we will end up.
And your point about what it means to “be rich” in a society where money doesn’t matter much, is a good one.
I should also say I am talking about the USA here or whatever country ends up hosting these AI companies. I’m sure there will be massive anti-trust actions taking place that might distribute it over more than one company, but certainly not internationally. Other countries will not have access to perform this sort of taxation or money printing, but will still be unable to compete with the AI countries. So they, including the UK, are royally screwed. I don’t know what is going to happen there but it is not good.
The doctrines that Fraser Orr is presenting in this thread (whether or not he believes in them, or is just presenting what others believe) are wrong – just wrong.
There is no point going over this ground over-and-over again. Various people, including myself, have already exposed the fallacies these doctrines are based upon.
@Fraser Orr: you still haven’t explained why the supply of goods is going to increase thanks to AI, to such an extent that all the people it puts out of work (and all those already not in work) can afford to live for free. And not only live for free, live lives of greater prosperity than today. All production of ‘stuff’ is backed by units of energy and raw materials. Will AI make that energy cheaper? Or produce more steel? Or get more oil/gas out of the ground? Or produce more food? No it won’t. Or at least if it does not by enough to be able to fund the lives of tens of millions of office dwellers who have been put out of work by AI. And all this is before explaining if those put out of work by AI are given lives of free leisure better than their old ones, why exactly are the oil and gas workers, high voltage electric cable engineers, sewage operatives etc etc etc are going to go out and work hard every day making sure the drones get everything they want? What exactly does society give them in return?
AI economics appears to be as follows:
1) Apply AI to everything!
2) Mumble mumble mumble…..
3) Everyone will live lives of plenty for free!
@Jim
@Fraser Orr: you still haven’t explained why the supply of goods is going to increase thanks to AI, to such an extent that all the people it puts out of work (and all those already not in work) can afford to live for free. And not only live for free, live lives of greater prosperity than today. All production of ‘stuff’ is backed by units of energy and raw materials. Will AI make that energy cheaper? Or produce more steel?
If you go back and read what I wrote you will see I was talking about digital goods and services. But you are right, some goods and services require raw materials, and certainly these AI companies require compute resources and electricity. But, again if you read what I wrote I talked about a few million dollars in compute generating billions of dollars in value. So, yeah, you are right to some degree, but it doesn’t change the reality of the massive deflation I am talking about.
All manufacturing is a process of converting cheap raw goods into higher value goods, whether that is transforming a rock into a sharp flint knife, or transforming some computers and a big building into a company that can ship you a steel knife overnight, with Prime Membership.
Is there a role for people in that extraction economy or other intermediate economies (such as steel production)? Maybe, though for sure many of their jobs would eventually be replaced by general purpose humanoid robots, or alternatively people in far flung countries who are screwed because the valuable goods are made for very low cost and all they can do is sell their stuff. And for sure, law is involved as a transaction cost in terms of property rights and intellectual property rights.
But digital and service goods are far and away the highest profit margin goods, which is why this all starts with them. As I pointed out originally, five of the world’s ten largest companies produce digital products only. Of the remaining ones, I explained how Amazon could be run by AI and some capital goods like robots and self driving cars, Three are actual manufacturers TSMC, Tesla and Broadcom though Broadcom may well be a design only firm, I don’t know much about their business, and the only resource extractor is Saudi Aramco.
So will there be some residual traditional business left? Initially lots of them, but over time less and less. So, it depends on the timeframe you are using to measure. But the high value stuff is all AI-able or (AI and robot)-able. It is why xAI is starting with Macrohard — a company to produce software and digital goods. We will have to see where it goes from there.
So, in summary the economics works like this:
1. Take some raw materials (a rock of flint or some electricity for example)
2. Some entity (either a flint knapper or an AI with robots) transform them into products and services more useful than the raw materials
3. This means that the value of the raw materials is less than the value of the products produced. The difference is profit.
4. Government taxes some of the profit, one way or another, and gives some of those taxes to people not involved, either as welfare or UBI or corrupt payoffs.
5. Some of the profit is given to the owners
6. Some of the profit is retained either for capital investment or to feed goods and services back into step 1.
At a high level it is pretty straightforward, no mumbling required. It is no different here than any other economic activity. The big difference is how much larger the profit leverage is in step 3.
@Paul Marks.
The doctrines that Fraser Orr is presenting in this thread (whether or not he believes in them, or is just presenting what others believe) are wrong – just wrong.
That is indistinguishable from me saying “The values Paul Marks espouses, whether he believes in them or not, that fiat currency and massive overspending are going to destroy the economy and result in terrible poverty, are wrong just wrong.” I know you are not advocating for this, but you are certainly saying that is what is going to happen. So, that is exactly the same situation when I am saying “this is going to happen, whether you think it is good or bad.” I think shooting the messenger went out with the Persian Empire.
I’m starting to understand the Butlerian Jihad more each day.
😉
@Fraser Orr: No an AI economy does not start with digital goods. It starts with real goods. Food, energy, housing, basic manufactures. The things people need to live. Digital goods are the icing on the cake, they aren’t the basis of an economy. Man can live without Grand Theft Auto 27, food clothes and a warm house, not so much. Explain to me how AI is going to make the basics of life freely available to all, and having done that how the jobs that can’t be done by AI are going to be remunerated. You are talking about an economy that is run without prices, and as far as I can see it’ll work about as well as communism managed without them.
An AI economy absolutely does start with digital goods and then over time it moves into other goods. These changes do not all happen at once, they take time. And to trivialize digital goods as “Grand Theft Auto” is disingenuous. A huge amount of the western economy consists of digital goods.
As to food, housing and energy, I suppose these will not be the first things to be moved into the AI economy, but they can obviously all be done in exactly the same way we do it today but with human brains replaced by digital brains and human hands replaced by robot hands. More likely they will be done considerably differently though in the long run. Perhaps you can explain to me why I would hire an electrician or plumber at $150 and hour to help build a home when I can use a robot to do exactly the same work and pay $30,000 for it one single time? It would make no economic sense. Of course one reason you might do it is because of unions or building codes, and that is that “human” friction I was talking about.
And where exactly did I say things would be free and there would be no prices? What exactly would the point of a UBI be if there was no need of an income?
This thread has become a waste of time.
With obvious fallacies, such as “Universal Basic Income”, and “money is just a means of keeping score” (i.e. not a store of value – just a medium of exchange) being presented as if they were serious ideas. Ditto utterly absurd ways of producing energy – being presented as if they were sensible.
And, of course, the claim that “AI” will produce a Star Trek, or Warhammer “Dark Age of Technology”, society where most people will not have to work.
The thread has become too stupid to bother with any more.
“And where exactly did I say things would be free and there would be no prices? What exactly would the point of a UBI be if there was no need of an income?”
You said that all the people displaced by AI would be granted a ‘generous’ UBI. Which in effect means they get to live for free. All the stuff they buy now, they’ll be able to buy with their UBI, which they just get given. No work required. And the greater the AI penetration of the economy becomes the greater the UBI will become. And you also accepted that once AI becomes ubiquitous that the concept of ‘money’ will lose its meaning. If no money, then no prices either. You still haven’t explained how when stuff becomes so cheap as to be meaningless why anyone will bother producing it, because they can’t sell it for anything because its virtually worthless.
And you also haven’t explained how we give all the people who lose their jobs to AI a ‘generous’ UBI, while expecting those who AI can’t replace right now (who are all the most important people in society basically as they ensure we have energy, fill the stores with food and build our houses etc) continue to work like slaves. Do you think its going to be politically acceptable to have a two tier society of Eloi and Morlocks? Free money for former HR managers, lawyers and accountants, while brickies, plumbers and care home workers are expected to get up at 7am and haul themselves off to work?
Here is an idea or a point about AI. AI as we know it now is a LLM – large language model. After training it “knows” what was contained in the TEXTS it used for training.
It does not learn, or make any contact with physical reality itself. It just learns from what other people (lots of them) have written. That is – it does not know anything new, cannot discover anything new.
So, most speculation above are just science fiction. It is fun, but not much more.
@Jacob
Here is an idea or a point about AI. AI as we know it now is a LLM – large language model. After training it “knows” what was contained in the TEXTS it used for training. It does not learn, or make any contact with physical reality itself. It just learns from what other people (lots of them) have written. That is – it does not know anything new, cannot discover anything new.
AI is great deal more than LLMs, and a great deal more than just the input texts. It is a common mistake to think that AIs only know what is put into them as source texts. On the contrary AIs do a great deal of experimentation on their own in a process called RL or reinforcement learning, which is a bit confusing since you’d think RL means “real life”, but it means exactly the opposite, AI’s running in simulated environments. Moreover, many AIs interact with the real world and frequently when inference does not produce good results then that is fed back into the training process. That’s why grok will often ask which of two answers do you prefer. And insofar as the RL training does not match the experience when applying to the real world, it adjusts the simulators to better reflect reality.
And, FWIW that is exactly how you learned. You took in some information from texts and direct instruction, but you also interacted with the real world, you experimented and over time updated your neural net with the results of that experimentation, either good or bad. This is true at multiple levels, it is how you know how to move your arm muscles to touch your nose and how you know how to ask a girl out on a date.
If you are interested in learning more I’d recommend Phil Beisel’s excellent and very accessible series of articles on how Tesla FSD works. I think you would find it illuminating.
https://x.com/pbeisel/status/1848807994894733530
Thanks for the link. I’ll check it. Meanwhile the are a lot of “Next Big Thing” claims mentioned there… When I see such claims I turn skeptical.
As to Tesla robotaxis — they have maybe a dozen experimental cars. Waymo regularly runs thousands of autonomous driverless cars, for more than a couple of years, in four or five cities in the US (and starting now in London), and have made already more than 100 million passenger-miles. And they don’t claim they use AI. (Neither that it is the Next Big Thing)
@Jacob
Meanwhile the are a lot of “Next Big Thing” claims mentioned there… When I see such claims I turn skeptical.
That’s definitely fair and a wise approach. Tesla’s first robotaxi came off their production line a couple of days ago, two and a half months ahead of schedule.
As to Tesla robotaxis — they have maybe a dozen experimental cars. Waymo regularly runs thousands of autonomous driverless cars, for more than a couple of years, in four or five cities in the US (and starting now in London), and have made already more than 100 million passenger-miles. And they don’t claim they use AI.
Tesla robotaxis will crush waymo for a several of reasons: first Waymo cars cost about $100,000 to produce, Tesla’s robotaxi in volume costs about $10-15k (this is mainly to do with vertical integration and the simpler sensors on Tesla cars and because Musk believes “the best part is no part”.) Second, although waymo may have 100 million miles, Tesla has about 8 billion miles on FSD because of course they have a vast fleet. Third, simple economics, in volume Tesla’s are planned to cost $0.25 per mile, Waymo is about $1.20 (I don’t remember the exact number but it is around there), and a regular self owned car is about $0.85 per mile. The economics simply do not work in Waymo (or the other companies’) favor. These numbers are of course in the US, cost per mile for self owned cars are much higher elsewhere because gas is more expensive elsewhere. Fourth, Waymo has cleared the way, that is to say they have set aside the regulatory obstacles in many places which makes it easier for Tesla to do business in those same cities. Fifth, beyond just Robotaxis, Tesla is planning to allow owners of regular Tesla’s to “loan” their car to the fleet with revenue share, which immediately makes it possible to add hundreds of thousands of cars to their fleet, including in smaller communities where it is less economical for a large corp like Waymo to set up.
Tesla robotaxi miles will be cheaper than self owned vehicle miles and I don’t doubt that that will lead to many people simply not owning a car and just using Robotaxis. FWIW, I would probably be one of them.
Waymo will stay in business for a long time I think, because they have bottomless pits of google money to support them, and Google will not want Tesla to completely own the market. Which is probably a good thing for tesla too since it’ll defer anti trust problems.
If I were to give investment advice, which I do not — only a fool would take the advice of some random dude on the net — but if I were I’d be buying a lot of Tesla and I’d be short selling Uber. I have no doubt that within ten to fifteen years Tesla will be the first company with a market cap over ten trillion dollars.
You see — you talk about what will be. I’m talking about what is. I’m not wise enough to know what will be. Waymo is. Tesla robotaxi will be. I’ll be able to talk about the robotaxi when it is.
Manifestly untrue 😉