Wouldn’t it be neat of a ton of really rich assholes, suddenly woke up poor?
Yes, but they’ll socialize the losses.
Oh no, I was a billionaire now I’m slumming it with 700 million…
More common: I am 65 years old and my retirement just imploded.
Eh they can’t let it get that bad or else they know humanity will truly embrace socialism and vote to strip mine the wealthy of their assets.
Why do you think folks like Peter Thiel and Larry Ellison are pushing “law enforcement” and mass surveillance so hard?
At least some of them believe they can credibly secure themselves against the masses easier than they can appease the masses.
I’ve been re-allocating my 401(K) investments for a few months now to try and soften the blow when it hits, but I’m not sure how far-reaching this will be.
I did the same thing. I moved to investments in the raw materials used to make the chips. I wanted to keep some AI related investments for the rise of the bubble so I moved my money out of microsoft, alphabet, amazon and nvidia and put it into Taiwan semiconductor manufacturing and Samsung electronics. The components will be already paid for while the software side of things implodes. At the beginning of Q4 I plan to move the profit and 50% of my base investment from component manufacturing into bonds so I can retain some of that value.
The AI bubble is going to fuck up the economy again. I had a ROI of 29% which seems insane. So many people are going to lose funds from their 401Ks when the bubble pops. Most people that have 401K through their jobs just let the entity managing the funds make the investments for them into something like retirement trusts/target date funds. Depending on the target date, most of what they are investing in is domestic funds where alphabet, amazon, meta are like the top stocks.
I urge everyone to read the prospectus documents attached to the trusts/target date funds you are investing in to see if you are in danger based on the top 10 stocks associated with your investment product.
Yeah, unfortunately my options for how to allocate investments are limited by our plan provider. I’ve been taking them out of any funds that focus on US tech stocks and opting for international.
That’s not gonna happen though. The rich will stay richer and the common pleb will be the ones carrying the weight of their failure.
If only this only impacted them.
The flow on effects of a major market crash are never isolated to just the stupid assholes building the bubble.
Be the change you want to see in the world
Anyone thinking this means the technology will disappear will be sorely disappointed.
Agreed, but maybe we can hope it won’t be crammed down our throats quite so hard as investors stop demanding such big returns?
A girl can dream, anyway…
Yeah it won’t go away but my job will mostly go back to normal. Right now they force us to use AI for everything even the stuff it sucks at. Even if I can do a better job manually they force me to use it. It’s insane.
I want to use it to help me with the stuff I suck at and let me do the stuff I am good at. I can’t wait for that to start. That’s how I have used it in my personal life and it’s been actually helpful. Not life changing though just helpful
It will collapse on its own overinflated expectations anyway. Its not the clean easy replacement for humans that CEOs are pretending it is.
But can it replace CEOs?
Please?
AI might not be able to, but the rest of us can take a stab at it.
That it won’t, but hopefully models will be more usable for private selfhosted use, and most definitely the related companies will go belly up which is a good thing and the datacenter boom will at least for the moment stop
It would also give that last push to the US to go over the cliff and end itself. I’m hoping it will kill the trump admin for good and make all politicians talk more about stopping the ultra rich for good
It’s like railroads, cars, radios, video games, computers, the internet .etc.
Any new world changing technology ends up making a bubble because investors live in the fantasy land of what could be possible. Eventually it pops and you get a more measured approach for what that tech is useful for.
The 1870s railroad crash didn’t kill railroad. Neither did bubbles popping in the 30s around cars or radios. Same for the internet in 2001.
No. They won’t take the hint. It’ll get a rebrand and they’ll do it all over again.
Pop already I can’t wait to see the cryptoturds and the prompturds cry
I’d agree, except it’ll ripple out to innocent regular people as well, just like the 2008 Great Recession, if not worse.
When regular people lose a lot of money, the richest get richer or are largely unaffected.
When the super rich lose enough money to actually hurt them, their interconnectedness with and control of everything that regular people need to survive means that millions if not billions of people will suffer immensely, up to and including dying because of it.
Is it selfish to think the collapse of the economy and potential loss of employment is worth it for component prices to drop back to normal?
deleted by creator
I mean… both are inevitable… the economy has collapsed for the normal people, and employments been hitting everyone for over a year, and bottom line, either way AI goes… it’s a lot of jobs. IE it fails and shrinks down… all the jobs around it are in trouble, it actually succeeds it’s going to take as many jobs as possible, there’s no job friendly ending to this.
I’m too burnt out to care about the economy or component prices anymore.
I just want the morons who got hooked on this shit and ruined their brains into mush using it to reap what they have sowed and face the music when their brain substitute costs more than they could ever afford.
They thougt they could game the system by saturating everything with metric cartloads of useless slop, the bastards, while those of us with still functioning brains waddled through it and attempted to fix it!? Well, fuck them, now they won’t be able to afford their crutch and their brain will be too far rotten to ever come back!
We’ll still have to clean their damn mess, of course, and the world will be much worse than before they and their suppliers ruined it, but at least there might finally be a light at the end of the tunnel.
I’m excited to see all of the AI slop YouTube channels die. They’ll have to use real photos for once. AI narration might end up being cheap enough that it sticks around unfortunately but at least photos and video will die.
I don’t think so. I sometimes think of covid and how when actual calamity hits we suddenly realize money is a kind of fiction. If the economy goes totally to shit we will see no mass starvation and probably not even mass homelessness or else we would have to admit what a giant failure capitalism is.
Just like covid, at the end of the day when shit hits the fan we retreat back to that archaic and ancient system known as “socialism”.
As opposed to the loss of employment driven directly by using AI that will continue if it doesn’t burst?
Ive been sitting on some cash waiting for things to tank so I can buy a new car below MSRP
It won’t disappear, but the business will drastically change, and the knock-on effects will be serious.
It will be like the dot com bubble in many ways. Back in the late 1990’s every other TV commercial was for Pets.com, Ask Jeeves regularly had a balloon in the Macys Thanksgiving parade, and Lycos was pushing to be the home page for everybody. They never recovered after the bubble burst, and companies like Google quickly took over.
In a similar vein I think a lot of the AI big hitters will vanish once this bubble bursts, leaving one or two lucky/cautious ones still around to scavenge the pieces. But I don’t think companies will be so hung ho about using AI for everything once that happens.
the internet was so exciting and interesting back then.
Compared to now where its just another corporate ruined hellscape that you cant even begin to get into without protection.
I could see generalist LLMs IE ChatGPT disappearing if only because they seem pretty fucken useless on average.
I don’t think they will. I’m the first to be massively sceptical of LLMs, but that doesn’t mean they can’t be used to build good tools. The key is recognising that tasks where correctness is vital should not be solved by the LLM directly. At my job, we’ve built an LLM-agent that’s very useful (internal use). What we’ve done is build essentially a Python library that this LLM uses to interact with our data. That way, we ensure that a query like “set up a skeleton for X” will be done correctly, while we save a bunch of time that would have otherwise been spent doing boilerplate work.
Basically: Enforce correctness by constraining how the LLM can interact with your data, and use the LLM to translate short natural-language queries into actions that otherwise would have taken 30 min of click-ops or write-run-toss scripting.
That’s exactly the opposite of what I’m talking about, the LLM that your work is using is while perhaps generalist for internal standards is still rather limited compared to the large more general models being such as ChatGPT or whatever the fuck musks model is called. No I’m talking about the fuck off massive models that have been scraping the entire internet in a vain attempt to create AGI, so the stupid cloud based models that have been being shoved into everything.
I think those are going to implode on themselves simply because they are too expensive versus the fuckall you get out of them.
My point is that we’ve built our model on top of these “generalist” models. You hook it up to an API and then let Claude, Mistral, etc. (I try to avoid GPT and some other) do the generalist job of translating human language into actionable tasks. You give it tools to parse documentation and actually do the tasks.
The generalist models are fairly good at taking a set of instructions and translating that to the correct tool calls, then our tools enforce correctness on the final output. Building an agent like the one we have would be nearly impossible without having some generalist model to do the “translation” step.
I think we’ll see two major changes going forward in how LLM’s are used: 1) they’ll become much more expensive and less widely used, since today they’re run at a loss. 2) they’ll be integrated into larger systems where they can do what they’re good at (parsing and outputting natural language), while offloading technical tasks to other tools that are actually built for technical tasks where formal correctness is paramount.
The two of you are using the word “generalist” differently. You don’t need your tool-using language model to be able to wax poetic about ancient egyptian burial practices. That’s why ChatGPT will become useless. It’s too large and expensive to continue running without subsidies, and it’s too useless for serious tasks. You can get away with a small local model that knows nothing about ancient egypt if all you need is to translate natural language into tool calls.
You’re absolutely right about that, and if we’re able to build a model that’s as capable as GPT and friends at parsing natural language, without simultaneously training it on everything from poetry to programming, that’s a major win. My current understanding of the field is that in order to build/train the models that are able to robustly parse natural language and “understand” the intent behind a series of instructions well enough to translate them to the correct tool calls, we need a very large and varied training set. I’m using “generalist” as a term to refer to the models that you can interact with in natural language across a wide variety of tasks. Those models are extremely powerful if you can also connect them to tools that solve problems deterministically, so that you get around the problem that they don’t really “understand” anything at all, while taking advantage of the fact that they’re extremely well suited for translating natural language to a selected set of pre-defined actions.
I think a major challenge going forward is that interpreting natural language requires a large set of training data. So training specialised models that can also interact with natural language is by nature difficult.
Get ready for GFC #2.
This one will be a lot worse than the sub-prime mortgage crisis, as AI investment has left many major banks, index funds, and most of the top 10 most valuable companies in the world heavily exposed, and the world governments are already at historic levels of debt - meaning a bail-out even if desired by those in power, may not be possible as it was in 2008.
Also there is more than 1 bubble that’s pretty much fully inflated right now: for example, Realestate.
Further, household debt has never been this high and is well beyond 2007.
Then on top if this there’s the record government debts in some countries (most notably, the US) and the weakening of trust in the USD thanks to a certain Mr D. Trump which might result in it losing its Reserve Currency status much faster, an event which would be massive, especially for the US Economy (it could very well trigger Hyperinflation).
When the AI bubble blows it will at the very least cause other bubbles to blow.
IMHO, this shit is going to be something of a level of the 2000 crash AND the 2008 Crash put together, possibly worse (especially in the US).
So on a scale of Great Depression to Dot-Com to 2008 Housing to Covid, where will this AI bubble burst land?
I would say it depends were in the World you are.
For one, the AI bubble (not necessarily AI as a whole, just the insane bubble around it) seem to be far bigger in the US than the rest of the World.
Also, I’m not so sure there’s quite as many lines of contagion between the US and the rest of the World for it. Granted, when the tide went out in 2008 it turned out there were things like German Landesbanken (regional banks) exposed to the subprime bubble in the US, so who knows who is actually exposed to the bubble in the US via indirect and more obscure channels such as loans to datacenter companies via the money markets.
My gut feeling is that in the US it will be close to Great Depression, in Europe it will be around the Dot-Com crash (possibly worse since most countries in Europe are deep in a housing bubble, though it depends of Europe’s exposure to the AI bubble in the US) and in places like China it will be more towards the Covid end of it (China’s investment in AI hasn’t been at all following the same Financial model as in the US).
That said, this is all a bit of guessing. I was in Tech in the DotCom crash and in Finance in the 2008 Crash so I like to think I’m a little better informed than average, but I’m still guesstimating:
The fundamentals look very bad (high consumer debt, overvalued USD juding by US prices, horrible P/E rations of most of the companies involved in the AI bubble, insane IPO valuations for the upcoming big AI IPOs and the whole shenanigans with indices that will push the losses to index funds and hence retirement funds - things which in turn might trigger a confidence crisis in US Markets - and so on), but who knows how tightly coupled things are or not with the rest of the Economy (especially outside the US, in the US there seems to be tight coupling with a lot of things via things like datacenter building and the impact of AI on corporate risk profiles and job markets), especially after the nasty experience some institutions had back in 2008 from things like recklessly exposing themselves to US Markets? It might very well turn out that by reducing trust in America Donald Trump was good for the rest of the World which won’t be quite as exposed to the US when shit hits the fan than they would otherwise.
Time will surely tell.
Let’s see how many governments will end up with a D rating. The USA are probably a given but this might set a lot of major companies on fire so who knows who else will run out of money.
Of course China is laughing all the way to the bank. Their economy isn’t super healthy right now but they aren’t reliant on semiconductor companies that chained themselves to the AI racket. So they might weather the crash mostly unharmed and we’ll all end up buying Loongson in the future because all of the x64 and ARM companies have folded.
Still time to inflate more. This season is still building up to the biggest and baddest. At the start of this season we hear names like Mythos and Glasswing but they don’t mean anything. Now we’ve learned they’re a new and faster way to discover vulnerabilities in software. The foreshadowing is building. We have the date, we know the upcoming catastrophe. In July, they will make public thousands of new software vulnerabilities. The internet will panic, software companies will spend billions on ai service to handle the damage. Anthropic will have a record IPO, followed by other AI companies. It’ll be YUGE. Stay tuned for the cliff hanger
Will it stop the invasion of data centers?
It should, no money to pay for the people building them does tend to make them not get built
Come on one more pump baby
I better hurry up and buy an appropriate motherboard to buy discounted RAM and CPU after the pop. Figuring on a Gigabyte TRX50 AI TOP. Dunno what the difference is between that and the 2B version. This MB is interesting, you can go regular Threadripper or the PRO, depending on what suits your wallet.
When is this burst gonna happen?
Knowing that would mean the ability to make lots of money off of it but nobody knows for sure.
Yeah it wasn’t a serious question.
Not soon enough.
Hurry! The government should buy 50% of its value like Bernie says.
My pronostic of Q1 2027 is looking as good as ever!
Let it brust, look what people do for money, power even to make their race first.
