It is 'nearly unavoidable' that AI will cause a financial crash within a decade, SEC head says
businessinsider.com
It is 'nearly unavoidable' that AI will cause a financial crash within a decade, SEC head says::undefined
It is 'nearly unavoidable' that AI will cause a financial crash within a decade, SEC head says::undefined
Is it because replacing employees with AI results in a never-ending cascade where your stupid system doesn't keep consuming because AI don't consume and won't get paid?
Or is it because using AI will result in the climate to continually become more inhospitable?
Maybe it will be because AI will be used to create more and more believable misinformation that results in WW3?
OK, it is addressed in the article...
He's specifically talking about the use of AI in finance, and that an algorithm that runs amok in a particular sector:
I'll throw out a microeconomic example. About a year into the pandemic, the price of used cars started going up... a LOT... in a short time. One of the reasons for the sudden changes in used car prices was that major used car resellers were using algorithms to set buying and selling prices for cars. While supply chain pressure on the new car market was unprecedented, and it trickled down to used cars, a facilitating cause is that the used car price-setting algorithms didn't really have any humans in the chain checking to see if the numbers they were kicking out made a lick of sense.
So you had companies like Carmax and Carvana buying used cars for $X, and then a month later 5X, then a month later 10X, because they were programmed to just up the offering price until they reached target stock levels. Sometimes they were buying 3+ year old used cars for more than the current price of NEW cars of similar trim level. Carvana's numbers got so whacked that it nearly sunk the company.
Now imagine that kind of a runaway algorithm in stocks, bonds, real estate, etc. It's 2008 all over again.
Honestly hoping something like this happens in residential real estate, if it isn’t happening already. Housing is well overdue for a correction.
You can’t tell me that most people can afford a $400,000-$700,000 mortgage. Median incomes don’t support that price point. Median household incomes might support the lower end…barely. So I am starting to wonder just who is buying/selling all these houses. When I see a $600,000 “average” house last 3 days on the market and then sell for $760,000…I have questions.
I swear if I ever marry it will just be to combine finances so we can actually buy a house and stuff
I am married with a household income in the 200,000-300,000 range and we can’t afford anything here.
I've been reading that nobody can afforrd to buy houses for at least a decade now and the price just keeps going up so clearly people can afford it.
In my blue collar, median wage earning workplace the vast majority are homeowners and having an investment property is seen as normal and expected., it's the new baseline for doing ok. They have dual incomes, two cars, and overseas holidays every year. They are migrants who had no bank of mom and dad and they prefer to send their kids to private or carholic schools.
They are not poor, but if you believe what you read on the internet they should have zero kids and be living paycheck to paycheck.
I bet every single one of them bought prior to the bubble and rate increase that started during the pandemic.
You are outta your mind if you think they are dropping $100,000-$300,000 for a down payment on a house, only to turn around and pay $3,000-$6,000/mo for a mortgage.
Those properties you see being sold are wealthy folks using them as investment vehicles.
Those numbers are very regional, though. $100K down payment is 20% for a $500K purchase, and lots of homes are under $500K. I just bought a new home (as in, literally just built & never lived in) here in California for less than that.
Not sure why people are downvoting your lived experience.
Investor purchases of single-family homes have spiked to 28% of home sales, and well above 30% in many high-demand localities. Market prices reflect that competition, so when one says, "the price just keeps going up so clearly people can afford it", one must also concede that many of those people are investors who are displacing buyers with less money. Those lower end buyers -- who could have afforded a house if an investor wasn't ready & waiting to flip houses or turn them into rentals -- have been frozen out of a home purchase.
I traded in a 2014 Toyota hatchback to Carmax and got an Audi A3 when the algorithms went haywire. It didn’t cover the whole cost but it was a silly enough trade that I thought for sure someone would call me and say it was a computer error.
Wow, what a deal! Did they say anything about how crazy that was?
No, they actually called and paid me $100 to make the swap at a lot about an hour away. I wasn’t gonna argue my way out of an upgrade so I was like, “Oh, yeah, I can drop it off wherever.” The dude who details the cars after you drop them off definitely wasn’t worried about it. He thought it was funny his bosses fucked up.
While that's really interesting, there was a lot more at play than a pricing algorithm. It was a culmination of a lot of things, starting with Cash for Clunkers that had a huge impact on the used car market. Then there were a ton of supply chain issues during COVID that squeezed the new car market. Probably some other factors I'm not aware of, too.
Right, my point is that relying on pricing algorithms when faced with novel "black swan" conditions nearly drove major used car dealers out of business. Obviously, the algorithm didn't cause the novel conditions, but neither did they buffer the effects. Instead, they accelerated the financial effects.
If you go back and read about Carvana before the pandemic, their big selling point to investors was that their algorithms were "smarter" than the competition and would realize more consistent profits for the company and their investors. When it became clear that these "smart" algorithms went insane, investors abandoned the company and their valuation dropped from $60 billion to $7.5 billion between 2021 and 2023. Carvana has narrowly avoided bankruptcy.
My 2013 Prius got totaled around the peak of this. I wanted to just replace it with the exact same model, because it's a good car. It would have been cheaper to buy brand new one at the time. I got a new electric car instead and with the $7k tax rebate ended up spending less than I would have to buy a 9 year old Prius.
He has a good point as this monoculture of systems and models would very greatly amplify any market imbalance and defects, at a speed human bank managers would only realize when getting notified of their impending bankruptcy
It was pretty bananas for a minute. The Mazda dealership offered us 5,000 more than we paid brand new for my wife’s Mazda 3 in 2018. I told the salesperson that it makes no fucking sense and he couldn’t explain it either. Didn’t go for it for a bunch of reasons but it was really odd.
My mom's 2020 Fit at the top trim level sold new for roughly 20k. Her lease buyout price was 1/2 the cost of an entry Fit in the same year with 30k miles (going for 25k at the time)
Currently, I would rather guess it's the usual bubble popping. AI has attracted billions of investments and will likely pull in even more, but it's already foreseeable, that hardly any of the investments will turn a profit. So we'll end up with a third dotcom bubble.
AI isn't a bubble. The futurist/Rationalist/transhumanist communities were saying what's happening now would happen in a few years about a decade ago, and our predictions are that the next phase is AI taking over all labor through sophisticated automation. We've been trying to warn everyone about this since the advent of Google Deep Dream, but sure stick your head in the sand again and let the world burn around you; it's worked so well so far.
This comment will probably get bombed, but w/e. 🤷♂️ Go ahead and be ignorant and angry at me, I'm used to it.
Edit: yes I am bitter. I've had a bad day, and I'm annoyed.
How many times has that been predicted already? Three, four? Look at the history of AI, it happens every few years.
Anyway, you're implying a dichotomy here. World domination or pipedream, but that's not the case. The dotcom bubble was without a doubt a bubble, but much of the underlying technology was used a few years later, just without the hype and fanfare.
AI will probably find its uses, and has the potential to eliminate a lot of jobs, but the current iteration of AI businesses is utter garbage. Even something as comparatively simple as Microsoft's Copilot is currently losing money - roughly as much as it costs to use. Yet, there are billions upon billions being poured into useless start-ups that will never produce anything of value in a profitable manner.
What exactly happened to self driving cars BTW? Weren't those totally on track of what experts predicted?
They're on the streets in San Francisco, I see them all the time
Yes. Definitely one of those or something else entirely.
That’s an economist level reply right there
I'm thinking yes, plus AI margin trading running into a tragedy of the commons where they collectively run the stock market into the ground and there's no reset button on that.
So do I collect my economic Bingo winnings after the 4th or 5th major crash of my adult lifetime?
You win the chance to afford to eat human or expired cat food you found on the body of someone you shot over a pair of worn boots.
Sounds like playing Zomboid.
I'm afraid your bingo winnings are going to be confiscated by the government in order to pay for more golden parachutes for the CEOs whose decisions led us here.
Bears have predicted 100 of the last 3 market crashes
Yeah.
It's a casino, and a lot of these folks are no smarter than a retiree rubbing troll dolls on their favorite slot machine.
The solution is to realign markets to human needs. The is no productive benefit to high speed trading. There is no benefit to stock by backs. We need to reign in most of the nonsense that these financial services institutions do.
This meme gets funnier and funnier every time I read it!
And then we finally eat the rich, right?
Sorry, the best we can do is a small marginal tax increase on income over $1M and a 50¢ bump to the minimum wage.
Better spread out that wage increase over a reasonable period of time, like 10-25 years. Wouldn't want to burden the precious job creators out there.
/s
50¢? Whoa now you arent trying to send us down the road of commies now are ya?
Better cut that raise in half at least
Well if we can't tax 'em we could always literally eat them, and if we can't afford food thanks to them killing the good paying jobs then maybe we will just have to eat them literally if we can't tax them properly.
But, you know, after the economy revives by state simping for the private banks, maybe we'll discuss it then
Fuck Gary Gensler. Dude tries to influence the market just to line his own pockets.
He should not be the head of the SEC.
Market manipulation? SEC should investigate him
Would you mind furnishing that statement with some links, articles, or some info to back it up? Iirc he's ex-Goldman Sachs but he seems to be making moves that hedge funds / market makers do NOT like. Vs the previous guy who just let them do whatever they wanted.
If you want to see the actual baddies, look at the 'self regulating' agency which has gone to great lengths to hide swap data, and the company which enforced 'position close only' on a large number of stocks (kicking off the infamous GME saga). Hint: it wasn't Robinhood. Robinhood was just one of many brokers who were instructed to set multiple stocks (not just GME) to 'position close only' aka no buying, only selling.
That is the job description of any regulator.
"Line your pockets with money from the people and entities you are supposed to reign in to protect the rest of the people who are subject to their actions so that your benefactors can prey on them."
They've been fucking with automated trading for decades though, unless you're going to try to convince me that there's a human investing in a trillionth of a company for a hundredth of a second at a time.
It's already caused "flash crashes" before. https://en.wikipedia.org/wiki/2010_flash_crash
The idea of investing in companies because you believe in them and want a share of their profits is sound enough I guess, but what it's morphed into is nonsense. The result is a system where you have trillion dollar companies that never actually turn a profit in favour of "growth".
The crash is already in motion and it was not AI that caused it. This is just getting us ready for who to blame when it crashes.
Many stocks have been far oversold and there is no way to reset it without the rich losing it all.
This is how they will reset it and protect the rich.
Is this before or after it destroys the economy for everyone but the super rich by replacing them and making them compete for fewer and fewer scraps? Sorry, there will be lots more new jobs created by AI probably, like AI wrangler, AI safety consultant and the like. Probably.
I always have a laughing fit whenever I see “Prompt engineer” used unironically in a job posting on LinkedIn.
I also think the term is granted way too much prestige, but a bit over a decade ago people also laughed at the notion of "Social Media Manager" being a real, high paying job
Who knows where this stuff will go
Even today the term Social media manager is still conflated with graphic designer, sales representative, customer information management, publishing, copywriter, photographer and creative writing, all the time.
I just applied for a Prompt Engineer job last week!
Honestly, it's a legit position. Maybe not something that will last a long time, but to do it well you need to know what prompts to give.
The average person might put "cat on a speed boat" whereas someone's job would need to know what "bokeh" is, what kind of camera lens you want to simulate with what F-stop, know the rule of thirds, negative space, etc.
The problem is whether someone is willing to pay for that extra knowledge or get their nephew to pump something out on their iPhone then say it's good enough.
I have... feelings about LLMs being the big thing in AI/ml right now.... because its really not much new. Maybe the transformer model kind of but ultimately LLMs are massive supervised learning neural nets trained on obscene amounts of data. And then other models use that pretrained "foundational model" to work and just tune their parameters. Which is why prompt engineer is becoming a thing.
Corpos are playing by the book here and trying to extinguish any competition before it begins by having people rely on their "foundation" models instead of innovating their own solutions
How many tutorials can you find for implementing LLM NLP tasks that dont include "import this model from X company" id wager its only maybe 33%
Part of what makes localized model engines and custom ML chips interesting is precisely their ability to enable small custom local models. Right now LLMs require so much computational power and massive amounts of data to be trained and operate that even the most expensive options lose money with every prompt query.
So, the reason every tutorial starts with "download this model". Is because there's a good chance you don't have the hundreds of super computer cluster chips and the several hundreds of exabytes of scrapped and curated data needed to train a natural language processing model. There's a reason there are only big players in this game.
Facts.
Even if you could design your own model... How do you acquire a dataset even a fraction of the size those pretrained models from the corps.
Then how do you train the model in a reasonable time. Other than relying on cloud computing which leads to the same problem of only corps can play this game properly right now.
I designed and collected/labeled the data for a relatively small deep CNN for my masters thesis and training it on 60000 images was taking over a dozen hours (this was 5 years ago at this point so that part may be misremembered) on a 1080ti.
Pretty sure that crash is more the fault of the greedy shits who think it's normal for 4 folks to own 50% of the country while 50% owns 2% of the country. Don't need AI to tell you that system isn't sustainable.
AI is just a convenient excuse and you know that.
Good, let it all collapse. I want to see the 1% shit themselves before climate change kills them.
That's not who suffers most in financial collapse.
Bingo. The super-rich love a financial collapse, it gives them a golden opportunity to turn disaster capitalist. All those foreclosed homes & businesses available at a knockdown price, nom nom nom.
I need to get off lemmy. Too many teenagers on here with their "burn it all down"s and "trust me, I have a super easy solution to a giant complex problem that has been plaguing humanity for generarions"s and "the bad guys are actually good guys, lol I'm so edgy"s
now now, no need to be dismissive of other age groups in this matter. I'm sure there's plenty of non-teenager people that think the same way too. on the internet, nobody knows you're a 74 yr old extremist
Is this suddenly a teenager platform whenever you want to infantilize the ones you disagree with?
No? It's...just always a teenager platform? It's not like the users change age from day to day.
Except for the fact that people literally do change age from day to day. Another day older.
Holy fuck, the level of immaturity and pedantry was even greater than I could have anticipated. Well done, you have out-teenagered yourself.
The only immaturity I see is the person throwing around ageist ad hominems in response to someone making a joke.
So, how do we engineer a situation in which the richest suffer most? End of capitalism?
Greatly increasing taxes for the super wealthy and closing tax loopholes would be a good start.
If they haven't done all this until now, what makes you think they'll do it then?
The rich will get off unscathed with a blank canvas to work on, and the poor will pay the price, just like with every other market crash.
Invent a system of money in which they are no longer in control of it.
Historically riots, mobs, and terror have been the answer. See the origins of the luddites, saboteurs, the French revolution(s), the socialist movement, etc.
The luddites failed and the French revolutions ended poorly for everyone. Not exactly the best examples to draw from if you are trying to encourage violent rebellion.
I was thinking more along the lines of mad max.
Financial expert predicts that (what is already) the longest bull market in world history will end within the next 10 years? And the thing that the world's largest companies are investing the most in might play a roll in that?
Bold.
Shock, awe, hysteria. More “AI” fear mongering headlines to boost the stock prices of tech giants. Yawn.
Lmao yeah. They need someone to blame for upcoming crash lol.
My thoughts exactly.
Heavy doubt on this one.
There is still so much misunderstanding on the state of AI and its potential based on current technology (spoiler: reduce your expectations significantly). How can you expect anyone to make predictions with such misunderstanding.
That said it kinda seems like a financial crash is already happening, regardless of AI.
I think it's tech illiterate people being amazed by chat gpt and shit, not realizing just how janky and limited it's actual "artificial intelligence"actually is.
I don't know why everybody keeps downplaying where AI is already at and the speed at which it is improving. It can already disrupt multiple industries with where image, voice, and LMM AI is at right now.
For me personally it's not that I want to downplay it, it's that I want to balance the scales. I see far more over-estimating of AI happening than downplaying.
The current form of AI is great as a tool and sadly there are definitely jobs out there that are nearly completely replaced by this tool. But that scope isn't about to change much based on where we are currently at. Many jobs require actual intelligence to make judgment calls, and the current form of AI just isn't going to cut it here as it has no real intelligence.
Of course, that won't stop dumb business leaders from still trying to use AI here, but that's an error in judgment that imo will correct itself over time.
The SEC means: Security, and Exchange Commission. In case anybody like myself hates abbreviations, and have to look it up on Google to see what it even means
Thanks, TIL.
Also, I don't even know why tf people downvote helpful answers like these. Bonus shit points if they also turn around and criticize Redditors for behaving a certain way. Ffs, man. /rant
I mean, a few communities I'm a part of have been warning about this since c. 2014, so I think he's actually correct in his prediction. I haven't read the article, but I don't think any solution he'd propose would be good regardless, so I think I'll just save my time. TLDR: failing a real leftist paradigm shift, we need global welfare like 5 to 10 years ago and UBI.
We need ubi. But that funding for a ubi program needs future predictions of economics.
And? Is it something this person thinks we would avoid without AI? You would need a lot of faith that 'the market' won't dunk on itself some other way. What a non-statement..
then maybe they should do something about it.... its not like they do not have time...
Have you seen congress? We don't do that here....
This is the best summary I could come up with:
The chair of the SEC has warned that AI could trigger a financial crisis, as Wall Street rushes to adopt the new technology.
Gary Gensler told the Financial Times that it was "nearly unavoidable" that AI would cause a financial crash as soon as the late 2020s or early 2030s, and said that reliance on models developed by tech companies could lead to economic chaos.
Wall Street banks have been enthusiastic adopters of generative AI since the splashy launch of ChatGPT last year.
Morgan Stanley launched an AI assistant based on OpenAI's GPT4 model to help employees access market information last month.
Rival JPMorgan, meanwhile, has reportedly filed a patent for an AI model known as 'IndexGPT' that would help traders choose securities to invest in.
The SEC did not immediately respond to a request for comment from Insider, made outside normal working hours.
The original article contains 326 words, the summary contains 144 words. Saved 56%. I'm a bot and I'm open source!
Yay
what doesn't cause a financial crash?
Weird shaped head.
The SEC is full of shit
They might be right on this one. It could be the next dot com bust unless some new buzzword comes along that everyone can latch on to.
perhaps, but they are wrong about a lot