The AI bubble is already popping; we just don't know it yet KETTLE Big tech's Q2 earnings have been pouring in over the last two weeks, and it looks like stock market investors are taking a pin to the AI bubble. You can listen to the latest episode of The Kettle right here on this page, as well as on Spotify, Apple Music, or YouTube, where you can subscribe to get notified about the …
Situational Awareness are in the shit because they didn't sufficiently understand they were operating in a risky field and guard against that?
It's almost as if they were lacking some particular form of insught, but I can't *quite* put my finger on what that might be called.
Quoting from an ABC Australia article today on the recent 40% correction on the S Korean market:
In the past five weeks, the market has collapsed, shedding almost 40 per cent, and threatening the livelihood of millions of South Koreans.
The writer contends that tech heavy (mainly Samsung SK Hynix) single stock leveraged Exchange Traded Funds† (ETFs) drove the exuberance and now doubts arising from the debt burden of those tech companies and the sector generally are causing serious misgivings if not panic… yet.
† I have no idea what single stock leveraged Exchange Traded Funds are but I assume they are ETFs with all your (nest)eggs in one basket.
Patrick Boyle has an excellent YouTube video on the problems South Korea is experiencing benefit of these ETFs.
https://youtu.be/nJtL9MBVj48?is=5hKX7DSrZzFlFlsI
Basically, they are so much of the country's economy that tracker funds have to buy them every time they rise slightly and sell them every time they fall slightly, resulting in wild swings and instability which are crushing the huge number of day traders and small investors.
most of the world is struggling anyway and the wars have exacerbated it. the global elite are insulated. the only people whose lives will meaningfully degrade will be middle class people in rich western democracies, who will still be incomparably more comfortable than the third world.
Have you visited the "third world" recently? A hell of a lot if people there are now better off than the working classes in the "first world". The terminology reflects the state of the world about half a century ago and quite a lot has changed since then, not least the disappearance of the "second world".
Hiiiiiii! I'm from the “third world”, otherwise nowadays known as “developing countries”, “post-colonial economies”, “Global South” etc. etc. etc.
I assure you, while we aren't in the cacky the way the Americans are right now (you lot are basically 50 small-to-mid-economies in a hegemonic and now totalitarian overcoat), or even the English (what with that decades-long own-goal), we assure you, things aren't exactly great over here.
Stagnant wages, rising energy prices, being squeezed between multiple nuclear-armed states… not great!
see, this is why i like being pseudonymous on the web. i am from the third world. thank you for educating me about my people. i’ll be sure to tell those that are lucky enough to have stable jobs that they’re better off than the working poor in the first world.
yes, most third world countries have a burgeoning middle class that has profited off of the globalisation of the past few decades and influx of dollar and euro salaries. this however only led to the further uplifting of sections of the populace who were reasonably well-read and educated and had inherent class privilege. the underclass in most of the third world is utterly invisible. very few expatriates to the west come from this layer of society. consider: who makes cheap garments in Bangladesh? definitely not those you are talking about!
in fact, the third world middle class enthusiastically participates in the further repression of the local underclass. their lives are more comfortable than yours because of the labour they steal from their own people.
So, the aircraft we are on is coming apart. But you think a soft landing is still possible? Perhaps you believe that in the brief time remaining to us,we can magically sprout wings and can somehow land safely or perhaps even proceed under our own power to our desired destination?
Maybe not impossible. But it seems damned unlikely.
A soft landing is indeed quite unlikely, but we still, at least as a society in total, have a decision about how bad the crash will be. Unfortunately, those of us who are aware of this and care aren't enough to make much of a difference with investors and far too much of a credulous public going the opposite way.
Not only do you misstate the situation as avoidable or unavoidable, ignoring the many different versions even though they all fall under unavoidable, but the comment to which you replied never said this was avoidable. They said, correctly, that even if you don't like LLM companies and investors and don't mind the damage they'll take when the bubble collapses, there will be a lot of other negatives which will likely be so severe that they outweigh the schadenfreude.
I have lived through Black Wednesday. I have lived through Black Monday. I have lived through Y2K. I have lived through the dot com bubble. I lived through 9/11. I have lived through the various wars the USA have decided to fight for no other reason than their own personal interests. I have lived through the credit crunch. I have lived through the end of the Mayan calendar. I have lived through the rapture. I have lived through two Trump presidencies.
All of which, were disasters the like we never seen. But see them we did. And we've seen all sides of them.
It won't be great. But we'll be fine. We were fine before, we'll be fine again. Just that little bit poorer. Again.
"All of which, were disasters the like we never seen."
No, most of those were not. I'm not sure the AI bubble will be one either. But many of the things on your list were comparatively tiny compared to quite recent history, so using them as evidence that we can survive anything is not useful. Also, nobody claimed we would not survive, so since planetary elimination wasn't anyone's expected outcome, pointing out that it won't happen isn't doing much to argue against them.
Buddy, starting out in the world as an adult in 2008 when your family are losing jobs left right and centre, going out trying to find work in an economy where companies were failing let alone not hiring, is not and was not trivial. People lost their homes overnight. People lost their jobs in an afternoon.
There was nothing trivial about an event that was considered to be the worst economic downturn since the depression in the 1920s.
It only feels trivial now because we got through it. It was not a trivial matter though. It's just now we have experience.
You said all of the things on your list were "disasters the like we never seen", and when countered, you successfully argue for one of them. One I didn't disagree with, either in my statement or in my mind, but if it helps, you're right, that one was a disaster that, while definitely not as bad as we'd ever seen, was at least bad enough that most of us had not directly experienced something similar. Let's take a trip to some of the other catastrophic disasters in recent memory that you made that claim about.
Y2K: The media story is that it was something everyone panicked about and basically nothing happened. As we know, the media story is wrong. In reality, it took a lot of work to make sure basically nothing happened. Work that, thanks to the panic, people actually did in time and paid for, so the disaster that happened was that some companies had to spend a lot on programmers familiar with old systems to fix a bug. That is not a disaster from anyone's perspective except companies that ended up going over budget.
The wars the US fought: The US has been fighting wars for a long time. There are ones before the 2000s with far larger casualty counts (civilian or military), with more economic consequences, with wider impacts on countries other than the US, with more effects on the US domestic populace. For that matter, so have other countries. What makes these wars unparalleled in comparison?
Then we have several financial crashes. The reason 2008 was so bad in comparison to these is that they caused a big spike in unemployment. The other financial crises you mention tended to have small effects if any at all, which limited the pain to investors. That didn't make them good, but they made them a lot less damaging to society. I can only hope the AI bubble will be the same. It won't be, though, so although it, too, won't be something the like of which we haven't seen, I predict its effects will be more like 2008 than 2000.
The claims hype and counter claims of AI's being worthwhile are pretty similar to tractors being adopted in the 1910 to 1940 period. The main difference is that it's happening at 5-10 times the speed.
Early tractors were heavy, prone to getting stuck in mud, difficult to start, and broke down often. Farm journals ran plenty of cartoons showing tractors bogged down or in pieces while a horse looked on smugly.
Not every technological development is beneficial. See the Sinclair C5, 3DTV, the Metaverse etc.
Tractors solved an actual problem. There was also a clear development path to meaningful improvement that made them viable.
LLMs (which is the “AI” in this context) are a solution still looking for a problem. There’s also no clear path that takes us from where we are towards genuinely useful AGI.
I agree in general, but although I don't like LLMs but they're not entirely without use. They're not good, but they're not blockchain.
They can be a great way of generating content-shaped data for your tests, they offer new ways of mapping the usage of language across large corpuses, in a business scenario they allow you to eliminate some needless tasks (using the heuristic that anything that can be done by an LLM does not need to be done.) There are uses for them, but they're are mostly weird edge-casey things and none of them come within a hundred billion dollars of these companies' valuations.
It doesn't impact your main point at all and they're clearly a dead-end that cannot bring us closer to AGI, but even cul-de-sacs can sometimes have their upsides.
The tractor analogy is a fairly good one, although I'd put modern LLM systems back at the traction engine level of tractor development. The development of engine power on farms is interesting in that for a very long time people thought in terms of "Engines power stuff, horses pull stuff" and stationary engines together with threshing machines got pulled from farm to farm by teams of horses. The conceptual leap of "Hey, why can't this engine propel itself from place to place" took quite a long time.
Steam traction engines were the first tractors and the first lorries, but neither were very good. It took several more innovations for the actual modern tractor to be developed; diesel engines, 3-point hitches and power take-offs that weren't just belt pulleys.
Artificial Intelligences are still at the "traction engine" stage of AI development, with other minor AI-like systems such as robot vacuums and robot lawn mowers skittering about at the edges (like early horse-drawn grass cutters, corn binders etc to follow the agricultural analogy). The idea is there, the concept is there but the implementation is sadly lacking.
We don't need machines that write code badly de novo each time someone asks for something. That's just daft and is also hideously inefficient just in terms of power used. Rather we need a building blocks method of enhanced software development whereby the AI helps assemble a Lego-blocks type of application following known patterns; as long as the building blocks are very rigorously tested and have strict guidelines for safety and security then this is a less CPU-intensive and less costly way to proceed.
We do not want to have humans interacting with LLM systems directly either; this seems to cause psychosis in both parties with the human generally coming off worse from the deal.
... is to live to see Jensen Huang - in his now totally shredded and grotty leather jacker - and Sam Altman huddled on the kerb, like the 2 guys in the Trading Places/Coming to America cross-over - but without someone shoving some cash in their faces, but only spitting in them.
Jensen will be fine. His wealth is in cash, based on physical products that have already been sold.
Altman's wealth is based on promises that he'll have a really great product one day, if only you give him enough cash today. Despite that, there's a good chance he'll be fine too because an IPO effectively passes the risk onto your pension fund and let's him cash out.
Perhaps I'm missing something, but you do understand that the point was that it's *meant* to be a transcript of a podcast consisting of an actual discussion/interview?
I mean, if you really want it in article format that badly, you *could* shove it into ChatGPT and get it to paraphrase the whole thing for you. Which would be sort of appropriate here.
Of course we know it. a) it's a bubble and so was inevitable, but more importantly, b) the panic rush to IPO by all the big names, not least the one that made Musk a trillionaire. Do you really think they'd be leaving money/growth on the table if they didn't predict that now is the peak? They're all cashing out before the crash, and dumping all the risk/liability onto the wider stock markets (including your pensions)
Leopold Aschenbrenner also worked for OpenAI (until fired) after FTX. That sort of person.
His book was completely invested in "AI" related stocks. No diversity.
His only "hedging" was shorting software companies - so basically exactly same bet as his long positions.
He was hugely leveraged - apparently 4x.
He didn't have liquidity in reserve to cover his leverage if his stocks dropped temporarily, hence the margin calls and the firesale.
Hardly a "good stock picker". It's easy to make a plus with leverage in a bull market...
The wolves of Wall Street saw this overconfident amateur with too much money (mostly other peoples of course) and took him to the cleaners at the first opportunity.
It was not about getting the best reviews for your models, getting chips / resources for new data centres or even getting revenue from customers. The big fight was for investor capital. Demand massively exceeded supply. Instead of filling all of them investors had to pick a toilet to flush their money down. They started to pay proper attention to how toilets work. We have reached the stage where the early investors are looking for greater fools. I am sure there are lots of them but the entire species does not have enough money between them to cover early enthusiasm.
That one has a propellant leak too. IPO price was $135. It quickly jumped to $210 and dropped to about $100. Currently at $108. That should be enough for another batch to cash out until the price drops to $100 again. At some point people are going to notice that SpaceX's promised total addressable market was based on orbital AI data centres with chips sourced from the moon. Can holders will be desparate to cash out at any price once the fools let that sink in.
Was it so they didn't carry the can for the investment bubble that paid for it?
quote: Mark Zuckerberg at Meta has figured they will build as much as they can now, and the AI will help target ads and help their core business enough that it'll be worth it,
It is hardly an act of genius to spend billions to continue doing what you have been doing for years.
GAFA have form here. People bought screens when they needed them, so they spent a bundle making 4K screens. People just bought screens when they needed them. Then they spent a bundle making 8K screens. People just carried on buying screens when they needed them. The market for screen fanboys was tiny. It was daft.
Hedging energy? If only the British government had hedged Sterling for the Brexit referendum! They would have made a bundle.
quote: is this entire quarter of earnings just an industry trying to pump air into this thing to stave off a collapse?
Yep. They will keep inflating, reducing their exposure, ensuring that the small print on their circular investments is solid. And then when the AI companies go down, they will buy up the IP in the firesale.
AI breaths money, eats RAM and poops lousy PR. Datasheds are not investments, they are subscriptions, paying for tech that has the lifespan of a hamster.
Sadly, the chatbots will survive, taking the easiest 5% of customer calls, remaining as an obstacle for most of us to get past when we need support. But that isn't what AI promised.
Politicians believed that AI would turn vape-addicted, skill-challenged youths into Master Chiefs to replace all those nasty foreigners they kicked out, and save the big hole they created in services and the economy. Not going to happen.
Most of GAFA will be fine. A bit less cash, but fine. But there will be losers, and a Lehman Brothers 2.0
What to do? Minimise your spending, retrench as much as you have to, ignore the shiny and the promises, and wait for the mania to pass.
There's a lot of wishful thinking going on surrounding AI. The foremost is that a sudden economic collapse will make people's lives better. The second is the thought that AI will save money by automating things people already do well. Third, the way to win the AI race is to thwart the competition by buying up all the GPUs, RAM, flash and electricity.
While the US Government did play a minor role in the cause - relaxing lending rules in the *1990s', and repealing Glass–Steagall - it was mainly greed and stupidity by banks (esp. synthetic CDOs) that caused the crash.
That no-one responsible went to jail (let alone was strung from the nearest lamppost) means the next crisis is pre-destined.
I believe that it was mainly in the USofA that the parties responsible for the 2008 crash did not face consequences. ..... LogicGate
Methinks, and believe it or believe it not, parties responsible for the 2008 crash and who did not face and suffer the consequences of their catastrophic failure have every very good and bad reason to rightly fear and be absolutely terrified about what AI can so easily do and may have meticulously planned for them specifically.
* .... And that’s not just an Alien Concept whenever a Universal Singularity.
There were risk analysts (often with an engineering background) at some investment banks who saw the danger at least a year before the crash.
They were, unsurprisingly, mostly ignored by their higher-ups. I believe Goldman Sachs may have been the exception, hence they're still around and firms like Lehman Bros. aren't.
Hmm, which bit don't you agree with?
That, for example, Lehman went bankrupt due to the 2008 crisis and Goldman Sachs didn't (and indeed made a profit by betting against mortgages, see https://en.wikipedia.org/wiki/Goldman_Sachs_controversies#2008_financial_crisis ) is pretty objective.
Specifically to Lehman, I'll add this anecdotal story out of interest: https://www.informath.org/apprise/a8800.htm
I only mentioned Lehman as an example, and Golman Sachs as the exception in that they apparently listened to their lowly risk analysts,
Methinks it be much more the case and virtual reality of the AI stock market investor circus of bubbles is already popping which is something altogether quite different from anything being of particular or peculiar concern to AI/AGI/LLMs and their Remote AIgents and Quantum Communication Control Systems .... AI@ITsWork.
* ...... but it appears as if y'all don't know it yet. Hmmmm? I wonder what is to blame for that?
A moving target that's still developing and might be shrinking ..... which is itself surely a quantum event where and when a this is also a that and together something else again altogether quite different ‽
And all might NOT be shrinking. Everything could be hardening to make AI Autonomous Intangible and Almightily Invisible which is yet again something else altogether quite different ..... and quite probably another quantum event very likely to be specific terrifying concern to more than just a very strange and/or ill self-chosen few.
Nah. At some point, the investors will see they are being scammed and the money will start flowing the other way as people dump the stock. At that point a full-blown AI winter will be upon us (again), and nobody will trust the concept (again) for a couple decades (again).
Or until someone dreams up a new name for it. Maybe they'll try Machine Learning (again). Etc. (again).
