Re: Wealth is a meaningless word
Except that the word “intelligence” never figured anywhere…
How the AI bubble could pop and take down the global economy, according to the BIS
The central bank for central banks is concerned about the eye-watering sums being invested into AI, and it's raising the specter of a global recession should the bubble burst. In its annual report for 2026, the Bank for International Settlements compared the current craze to historical events, including canal and British …
-
-
-
-
Friday 3rd July 2026 08:12 GMT Rattus
Re: Wealth is a meaningless word
the problem with "like brexit then" was I've already been screwed by dot com, 2008 financial crash, brexit...
I am investing heavily on KY and other loube technologies at the moment because I know I am going to be shafted again and may as well try to make it less painful
-
Tuesday 30th June 2026 15:51 GMT Elongated Muskrat
Re: Wealth is a meaningless word
So, business as usual, then? My wages haven't kept pace with inflation for years, and I doubt many other people's have, either. All the money is being soaked up by private capital, becoming ever more concentrated in fewer hands, and to be honest, if those with all the capital take a hit, so be it; it's their turn
-
Tuesday 30th June 2026 22:28 GMT powershift
Re: Wealth is a meaningless word
I'm sure it will be worse in America than anywhere else. Look back at how Trump handled covid, by throwing it under the carpet for 4 months until there were so many dead bodies they had to act to save their tax income.
Same approach is taken every other problem like our economy. One example is firing people that don't lie enough about jobs reports. Well all the shit he is throwing under the rug will surface when the bubble bursts. He said himself, "Everything I touch turns to gold." So yeah, it may be global problem but America will be ground zero.
-
-
Tuesday 30th June 2026 06:04 GMT Anonymous Coward
Re: Wealth is a meaningless word
All that will happen is that investors will become more hard-nosed: requiring proof of claims, demanding real-world measurable advantages and stop throwing (other peoples) money at every proposal that has the letters A and I in it's title.
They will only be hard-nosed until the next fancy idea comes along they cannot see the full implications of and then it's back to the FOMO races with whatever capital they can scrape together or leverage off assets (usually companies whose value then will get destroyed in the following crash).
What annoys me most is the desire to control the whole market, to become yet another monopoly. It's really not the only operating model out there. It's pure, undiluted greed.
-
Tuesday 30th June 2026 08:29 GMT Pete 2
Re: Wealth is a meaningless word
> the next fancy idea comes along
Quite. One of the guiding tenets of investment advisors is never be wrong alone. As long as all the other "experts" are saying buy, buy, buy! then it doesn't matter (to them) if they are all wrong. After all, it's not their money they are gambling away.
It is only the minority dissenting voices who ever get criticised when they are wrong. When they are right, nobody acknowledges it.
-
-
-
-
-
Tuesday 30th June 2026 06:08 GMT Anonymous Coward
The problem is that it never hits the people who cause it but innocents. If it would just hit the idiots whose existence seems to be based on lurching from market abuse and disaster to market abuse and disaster I couldn't care less, but the vast quantity of victims they create in the process is not OK.
-
Wednesday 1st July 2026 12:57 GMT Anonymous Coward
Reply to me as if I'm five, and be gentle.
Why will this not just affect the AI firms (and hardware data centers etc.) and the investors? It's not like the housing crash, which affected many people.
Sure there will be less private investment for a while, but it's not like there will need to be government bailouts. Rich people lose their yachts.
Have the pension funds really invested all out in AI?
-
Wednesday 1st July 2026 15:18 GMT cyberdemon
Re: Reply to me as if I'm five, and be gentle.
I am not a Finance guy, but Yes, pension funds have invested in AI in a number of ways..
Firstly "Index Funds" which invest in (as the link says) "a broad range of assets that mirror a specific stock market index, like the UK FTSE 100" or er, NASDAQ. That means your pension fund will automatically buy shares in companies like Oracle, SpaceX (aka xAI), and CoreWeave, alongside Microsoft, Meta, Amazon, Alphabet (aka Google) etc. And OpenAI and Anthropic too if they are ever allowed to go public.
Secondly, these companies aren't just going to let down their shareholders - they also issue bonds, which are "IOUs" supposed to be repaid in a fixed time e.g. 10 years. These IOUs can be sold onwards though, so their "value" (determined by what someone might pay for an IOU from Oracle, to pay 100 million dollars in 10 years time) can diminish (because of the changing estimates of probability that Oracle will still be around in 10 years and able to pay up). Bonds are bought by institutions - as my second link says: "HSBC, Bank of America, Citigroup, Deutsche Bank, Goldman Sachs and JP Morgan were active bookrunners on the latest [Oracle, $25Bn] bond deal.". These are "high street banks" which operate people's personal bank accounts, savings and pension funds.
If the $25Bn of bonds that Oracle sold later turn out to be worth as much as used toilet paper, these banks are in big trouble. And it is looking dodgy, as these bondholders (i.e. banks) have already sued Oracle, for causing the value of those bonds to drop, because they have wasted so much money on AI that "the markets" (i.e. people who DeutscheBank or HSBC might want to sell their Oracle bonds to) no longer believe that Oracle will be able to pay up, and so aren't buying them. Suing the issuer of your bonds seems like a pretty extreme step to me, because it will naturally decrease the ability of Oracle to be able to pay back the bonds, which means they could lose even more of their value. But maybe it is a way to push ahead of the shareholders in the queue to get the money out of Oracle when it eventually implodes.
And thirdly - the real reason for the 2008 crash wasn't just that banks had lost money on subprime mortgages, it was because those banks were in "leveraged positions", meaning that they had lent out way more money than they actually held in deposits. This is legal, but there is a minimum ratio of "real money" that a bank must hold. Since 2008 this "leverage ratio" was tightened in most places, meaning that banks must have a higher proportion of "real money" compared to "money that is supposed to be paid to them in the future", but some of these limits were relaxed last year, meaning that banks can once again put themselves in 2008-like precarious lending positions.
Stocks (as far as I am aware) are treated as "assets", so if a bank holds a bunch of tech stocks and their values plummet, then they can suddenly find themselves at the leverage ratio limit at which point the regulator may step in and tell them to suspend trading.
As I say, I am not an economist or accountant, but I did read a great book in 2010 called "Wh00ps! Or Why Everyone Owes Everyone and No One Can Pay" by John Lanchester. I highly recommend it to any gifted five-year-old ;)
-
-
-
Tuesday 30th June 2026 01:59 GMT Gary Stewart
To be...
While there are some very important applications that AI is being successfully used for it is being grossly over hyped in many others. I guess we need to wait for AI++. Given it's disastrous effects on memory and SSD prices, power requirements, and water usage I would really like to see a return to reality/sanity in the very near future. Say is that a pig on the wing I see?
-
Tuesday 30th June 2026 09:08 GMT rg287
Re: To be...
While there are some very important applications that AI is being successfully used for it is being grossly over hyped in many others.
Machine and Statistical Learning are being successfully used for some very important applications.
The stochastic parrots are not. Don't need them, don't want them and the frontier researchers like AMI have already abandoned them as a dead end in favour of World Models.
-
-
-
Tuesday 30th June 2026 15:58 GMT Anonymous Coward
Like Cassandra, the thing about autistic pattern recognition, is that what is blindingly obvious, and inevitable to us, is cryptic and occult to the neurotypicals, and they are the majority. We can sound the alarm all we like, we will be ignored, and then, after the fallout, probably blamed for not sounding the alarm. Nobody likes people who are right, especially when they are wrong.
My advice is to foster skills that don't rely on AI, and to stay out of the blast radius.
-
Wednesday 1st July 2026 13:06 GMT Anonymous Coward
I recently had it on good authority that a lot of serious industry folk lurk here on the Register forums, and that companies are routinely scared about how both the register and the commenters will pull apart any PR or hype in their press releases, and how we react to IT news in general.
Apparently we have quite a reputation and are well regarded.
-
-
Saturday 4th July 2026 21:34 GMT Anonymous Coward
Yes! We have a reputation for sarcasm and cynicism, and it's the cynicism and technological knowledge without bullshit when we pull apart some new technology (whether positively or negatively) that does it.
It seems that wading through the in-jokes, awful puns, and some of the more unusual posters is either part of the appeal, or more likely, the price worth paying!
You may post sarcastic comments, but when the topic is robot manufacturing or concurrent systems design, your knowledge in your posts will come through, and that's the reputation we have.
-
-
-
-
-
Tuesday 30th June 2026 04:54 GMT Richard 15
I think the phrase was irrational exuberance.
The reality is, there will be consolidation and there will be at some companies that will go belly up.
AI is getting a lot of "free" users, but sooner or later people need to pay and lots of people cant.
I have no idea how many AI companies will die, but I think we'll see a number of companies setting up AIs the way
they currently set up networks and file servers. A lot of people will have personal AIs once they start ramping up
enough memory production. Honestly, some relatively recent game computers that are considered "slow" will
likely be more than sufficient for personal AI machines.
The only question is how will they get trained/updated.
HR will likely get specialized AIs and they can be sold quarterly updates.
Law firms will pay for localized copies that the equivalent of all those law books.
Search engines will get replaced and jobs like search engine optimization will likely die out for the most part.
The good news is that AI is not really actually intelligent yet. Its a great simulation,but it can't replace competent
people for complex tasks if for no other reason that people don't know how to ask the right questions yet.
-
Tuesday 30th June 2026 08:49 GMT Dr Dan Holdsworth
Re: I think the phrase was irrational exuberance.
Current AI is just auto-correct on steroids, trained on huge bodies of text and using internal statistical weighting engines to decide what should follow what. That's how the training works; you are simply getting better and better auto-correct the bigger the model and the greater the amount of text input.
Actual intelligences try to make a model of the world, or work on simple instinctive models that are then refined. The nematode worms I did my PhD work on had a simple chemotaxis model: as long as the smell intensity stays the same or increases in strength go forwards, when it decreases turn in circles. That is really all that they needed in their world. Higher animals build models of how they think the world works and react according to sensory input tied into the model.
Actual intelligence seems vastly more efficient in energy terms than does LLM AI, and also seems vastly better at dealing with things outside the scope of the training.
-
-
Tuesday 30th June 2026 05:45 GMT TReko
Early days
What I find odd about the whole LLM craze is that the technology is immature. The attention mechanism it is based on, was discovered to be useful by researchers working on translation. LLMs came about almost by accident.
Who knows if there isn't another algorithm lurking that requires a fraction of the processing power? It's like everyone is going all-in on ISDN lines while fibre might be discovered next year.
The historical analogy is the canal building phase of the industrial revolution in the early 1800's. Soon these canal builders went under because of railways.
-
-
Tuesday 30th June 2026 08:45 GMT Steve Davies 3
Re: Canal Building
the workforce who had been building them transitioned into Railway Navvies. Many of the same skills applied.
I don't see this happening with AI. All it seems to me is that there are millions of fewer workers needed which will in turn reduce purchasing power and company bottom lines will suffer as more and more people live in abject poverty. Meanwhile the likes of Bezos, Zuck and Musk get richer and richer. May they all end up sucking on this ----> [see icon]
-
Tuesday 30th June 2026 08:55 GMT Dr Dan Holdsworth
Re: Early days
The railway-building mania burned a lot of speculators but mostly didn't hurt the actual companies doing the railway work, because they had actual income. When the AI bubble bursts, a lot of big companies are going to get lumped with an awful lot of kit on their hands plummeting in value due to simply not being cost effective to keep running.
-
Tuesday 30th June 2026 09:24 GMT Peter Gathercole
Re: Early days
But the railway bubble, like the canals, left a lot of long-lived infrastructure, like cuttings, tunnels, track beds, bridges, marshalling yards, stations et. al. which then helped drive industrialisation (and much of it still used!) even if the companies that built it no longer existed.
One really wonders whether bit-barns full of high maintenance, short-lived, specialised hardware that can't be easily re-deployed will have the same impact in the years that follow any AI downturn.
One of my concerns is that if someone does find a way of re-using some of this potentially redundant kit, it's going to have knock on effects on the further supply of new systems in an already shrunken market, causing the technology suppliers to leave, leading to a medium-to-long term supply problem of new systems.
-
-
-
Tuesday 30th June 2026 07:17 GMT SnailFerrous
Compared with some prior bubbles, we won't even get anything long term out of this. As with the infrastructure for the dotcom boom, most of the investment will depreciate to zero within a couple of years, with the possible exception of the data centre shells, electrical and cooling systems.
Many British canal and railway mania investors lost their shirts, but we still have and use the canals and railways that were built to this day, two centuries later.
-
Tuesday 30th June 2026 22:42 GMT cyberdemon
At least canals and railways were still useful centuries later.. AI bitbarns are nothing but energy guzzling water evaporators.. If as it turns out, they can't produce anything of more value than the energy they consume, then they are no use to man mor beast and will be dismantled for scrap copper. Not even the memory chips can be reused.
