Is OpenAI Becoming 'Too Big to Fail'? (msn.com)
- Reference: 0179942580
- News link: https://slashdot.org/story/25/11/02/1945203/is-openai-becoming-too-big-to-fail
- Source link: https://www.msn.com/en-us/health/other/is-openai-becoming-too-big-to-fail/ar-AA1PEJuE
"Its future is uncertain beyond the hope of ushering in a godlike artificial intelligence that might help cure cancer and transform work and life as we know it. Still, it is brimming with hope and excitement.
"But what if OpenAI fails?"
> There's real concern that through many [2]complicated and murky tech deals aimed at bolstering OpenAI's finances, the startup has become too big to fail. Or, put another way, if the hype and hope around Chief Executive Sam Altman's vision of the AI future fails to materialize, it could create systemic risk to the part of the U.S. economy likely keeping us out of recession.
>
> That's rarefied air, especially for a startup. Few worried about what would happen if Pets.com failed in the dot-com boom. We saw in 2008-09 with the bank rescues and the Chrysler and General Motors bailouts what happens in the U.S. when certain companies become too big to fail...
>
> [A]fter a lengthy effort to reorganize itself, OpenAI announced moves that will allow it to have a simpler corporate structure. This will help it to raise money from private investors and, presumably, become a publicly traded company one day. Already, some are talking about how OpenAI might be the first trillion-dollar initial public offering... Nobody is saying OpenAI is dabbling in anything like liar loans or subprime mortgages. But the startup is engaging in complex deals with the key tech-industry pillars, the sorts of companies making the guts of the AI computing revolution, such as chips and Ethernet cables. Those companies, including Nvidia and Oracle, are partnering with OpenAI, which in turn is committing to make big purchases in coming years as part of its growth ambitions.
>
> Supporters would argue it is just savvy dealmaking. A company like Nvidia, for example, is putting money into a market-making startup while OpenAI is using the lofty value of its private equity to acquire physical assets... They're rooting for OpenAI as a once-in-a-generational chance to unseat the winners of the last tech cycles. After all, for some, OpenAI is the next Apple, Facebook, Google and Tesla wrapped up in one. It is akin to a company with limitless potential to disrupt the smartphone market, create its own social-media network, replace the search engine, usher in a robot future and reshape nearly every business and industry.... To others, however, OpenAI is something akin to tulip mania, the harbinger of the Great Depression, or the [3]next dot-com bubble . Or worse, they see, a jobs killer and mad scientist intent on making Frankenstein.
>
> But that's counting on OpenAI's success.
[1] https://www.msn.com/en-us/health/other/is-openai-becoming-too-big-to-fail/ar-AA1PEJuE
[2] https://www.wsj.com/tech/ai/is-the-flurry-of-circular-ai-deals-a-win-winor-sign-of-a-bubble-8a2d70c5
[3] https://www.wsj.com/tech/ai/big-tech-is-spending-more-than-ever-on-ai-and-its-still-not-enough-f2398cfe
AI is a fraud until they get the I(intelligence) (Score:2)
Today, all AI companies are just selling fancy automation. While praying they are the ones who make the big discovery that delivers the I.
Re: (Score:2)
And nothing is too big to fail - not even Microsoft even though they try their best to hold our computers ransomed by Bitlocker and the Microsoft accounts.
Many corporations should really start to think about this when they activate Bitlocker and starts to have logins through Entra. One day those things might be inaccessible and your data is nowhere to be found. Do you have a local backup and an emergency plan?
Re: (Score:1)
You do realize you can print out your bitlocker key right ?
Re: (Score:2)
And when I need it that key has for some reason been re-generated.
This happens quite often at my workplace, not sure why, but maybe it's at every BIOS update when the Bitlocker gets disabled and then re-enabled after the BIOS update. So don't depend on it.
Re: (Score:2)
> This happens quite often at my workplace...
There are multiple ways for companies to centrally manage bitlocker keys, either on-prem or via 365. If you run into this frequently at your workplace then it's a skill issue with your IT dept.
Re: (Score:2)
Counter point: Do you need automation or intelligence for the daily tasks? "Intelligence" is a nice lab experiment and will have a lot of philosophical implications, but for all current use-cases you need good working LLM and image generators. You neither need intelligence for a reliable knowledge model (in particular knowing what it doesn't know), nor for coding help, image generation, or other automation. You just need models that work well. I bet there are also a few more interesting architectures to be
Re: (Score:2)
We don't have reliable knowledge models built on LLM's. That's the core problem which they're unable to solve, still. There exists no LLM based knowledge model which works well, so that's still science fiction.
LLM's have use cases, sure. But they're a lot less general than they're being sold as.
Re: (Score:2)
There it is!
ah, get off them already (Score:1)
They’re trying to do something genuinely useful for everyone. It’s not really their fault that the markets are so eager for growth that expectations, and the money pouring in, are way over the top. At some point, there’ll be a correction and a lot of investors will probably be disappointed, but honestly, that’s just how capitalism works.
If all of AI went away today (Score:3)
A lot of investors would lose their money, and a lot of students would mourn the loss of their homework "help", and a lot of people would have to work a little harder to find answers to their questions (like having to actually look at the Wikipedia page). But too big to fail? No, not hardly. OpenAI is *not* that deeply embedded in our lives. We still remember how to do things without AI, I mean, it's been maybe 3-4 years now?
If OpenAI were to fail, would a big bubble pop? Yes, I think so. Would we all be doomed? No, hardly.
Re: (Score:3)
> We still remember how to do things without AI, I mean, it's been maybe 3-4 years now?
I asked chatgpt how long it has been part of culture, and it confidently said it's been 26 wonderful years.
Re: (Score:2)
>> We still remember how to do things without AI, I mean, it's been maybe 3-4 years now?
> I asked chatgpt how long it has been part of culture, and it confidently said it's been 26 wonderful years.
At this rate that will become indistinguishable from truth in another 3-4.
Re: (Score:2)
The question isn't whether we can do without OpenAI's products: as you point out, it would be an inconvenience but not critical. The real risk is that with all those huge financial deals, when OpenAI fails, they'll take other companies with them, companies that produce goods and services that we cannot easily do without.
Re: (Score:2)
Yes, I agree. That was the part about the investors. Investors have plowed money into OpenAI that they *knew* was at risk. When a big bubble pops, it does take some innocent bystanders down with it. But that's not the same as "too big to fail" which would take down the whole economy.
Re: (Score:2)
And, unlike last time, there really *is* no money to save this industry. Just sayin', the Social Security Trust Fund will be in the red no later than 2030. A lot more people will care about that than AI when that arrives at our doorsteps ... not that we have not known for decades.
So yes, the bubble will pop, the useful part of "AI" will survive, the U.S. will be in recession, we all loose a shiny and "free" toy.
Moving on ...
No (Score:3)
OpenAI technology is still, while used at large scale, mostly experimental. They also do not seem to have anything g that is years ahead of their competition, wherever they are today is at best a few months ahead of their competition.
They do not really have a moat, and it is trivial for customers to switch.
"Too big to fail" doesn't mean "bubble too big" (Score:2)
"Too big to fail" refers to big banks or businesses that are so big and so embedded in our lives, that if they were gone, our economy would literally unravel. Chase Bank comes to mind. OpenAI isn't even in the same league, in terms of impact, should it fail. Yeah, it would hurt, but life would go on.
Re: (Score:2)
> "Too big to fail" refers to big banks or businesses that are so big and so embedded in our lives, that if they were gone, our economy would literally unravel. Chase Bank comes to mind. OpenAI isn't even in the same league, in terms of impact, should it fail. Yeah, it would hurt, but life would go on.
Also GM and Chrysler, as mentioned in the article, employ a lot of people. OpenAI employs relatively few, and if they are really good at what they do should employ less and less as time goes on.
Re: (Score:3)
Currently, think that is true. However, there are analyses that the stock market and the economy is only being propped up by AI. Were OpenAI to fail, it could cause a cascade. That could get serious quickly. I don't think we're at that point yet, but the more the C-Suite banks on AI, the more of a problem the U.S. will have when the bubble pops.
Re: (Score:2)
I think there needs to be another thing, a market retraction by itself I would say is not deserving of the status. That's what I would call it also, it's not a collapse forever, it's a retraction. Yes it will suck for awhile but besides the inflated market values AI is not that valuable and it's incredibly fungible, there's no lack of options that do similar things at similar qualities.
Also a point against is that the fact that if OpenAI goes out of business other than the financials the whole thing will k
Re: "Too big to fail" doesn't mean "bubble too big (Score:2)
Other than people buying chips specifically for AI nothing is really dependent on it though. Sure, there's a lot of Execs telling investors things about AI "adoption" but that's mostly just bullshit so that investors feel reassured that company X isn't missing out on the next big thing. And for the few who actually are using it to do actual work, losing Open AI itself is of no particular consequence.
Re: (Score:2)
> Currently, think that is true. However, there are analyses that the stock market and the economy is only being propped up by AI. Were OpenAI to fail, it could cause a cascade. That could get serious quickly. I don't think we're at that point yet, but the more the C-Suite banks on AI, the more of a problem the U.S. will have when the bubble pops.
Are you saying the entire American economy is lifting itself up by its own bootstraps?
Re: (Score:2)
Yeah, how much of people's retirement savings are invested in AI companies through various funds? I would say a fair amount. Not to mention most of the growth in the market for years has been in these ai companies.
That doesn't make them too big to fail I think, but it does mean that when they fail it will be a pretty big disaster.
Years worth of investment that could have been poured into useful (and valuable) things will have just been spent on decaying, unused data centers.
Years worth of intellectual inves
Too big to fail (Score:2)
Means letting them go takes down the rest of the us economy. They're not a bank and only have 8k employees.
They can go by by and it won't matter in the slightest.
Now the fact that we're all basically held hostage by the banks is something we ought to deal with but we a scared of socialisms so that ain't happening
I look forward to ... (Score:2)
... dumpster loads of GPUs hitting the used parts market.
Re: (Score:2)
I think the idea is this is a huge bubble and a large number or all of the AI companies fail - its hard to imagine they won't have vast quantities of GPUs they suddenly can't use. They can try to sell them to the surviving ai behemoths, yes, but it seems like a stretch to imagine they'll just want to buy all the old gpus of their former competitors all at once. Yeah, a lot of these things aren't gpus, which probably means the bankrupt companies will be more incentivised to sell the fungable GPUs rather than
Whenever this is asked (Score:3)
> when certain companies become too big to fail
If you have to ask the question, then "too big" has failed.
You've got a completely broken system whereby big for-profit corporation root themselves to deep that they then demand financial help from the public/government whenever they can't make enough money, even though they're supposedly valued in the $trillion.
free market american hypocrisy (Score:1)
nothing that has ever been called "too big to fail" has ever justified the name, and we have only delayed the inevitable and propped up the white-collar criminals who brought them to the brink in preventing their deserved and self-inflicted collapse.
The only actual solution is to create a government strong and competent enough create a reliable social floor SO THAT THE FREE MARKET CAN OCCUR
but go ahead and give more handouts to the rich. I mean you handed the biggest grifter on the planet the keys to the n
If these big businesses truly depend on OpenAI... (Score:2)
OpenAI can increase their rates if they need more money, and their customer businesses can decide whether or not OpenAI is worth the increased fees.
Unlike the 2008 situation, people aren't going to lose their homes if OpenAI fails. And, frankly, they're not gonna lose their jobs - businesses have been using "AI" as an excuse for cutting jobs. In fact, one could reasonably argue that OpenAI going under might result in an increase in employment!
Altman: whether we burn $50b a year, I don't care (Score:2)
What makes you think this guy cares about too big to fail or paying back investors? ...May 3, 2024 "Whether we burn $500 million a year or $5 billion—or $50 billion a year—I don't care, I genuinely don't," he continued.
"2% of Amazon.com’s sales" (Score:2)
Amazon didn't post a profit for the first 6 years, and OpenAI is still working on it's for-profit status. They reportedly have more than 700 million weekly users and climbing with a sharp trend upwards this year, so I don't think there's anything for them to worry about. Like so many things in tech, this is foremost a battle for market share. Being profitable ultimately depends on winning that fight.
Personally I don't care much if OpenAI succeeds. I use Anthropic products, they are much better for my purpos
Re: (Score:2)
This. I'm so sick of the "They haven't made a profit..." tripe. They are in their growth phase in a rapidly growing and evolving market. None of their investors expect them to be making a profit right now. They could make a profit tomorrow if they stopped reinvesting their revenue in building out the company and started to coast. Of course they would be quickly outpaced by their competitors.
Re: (Score:2)
I dislike OpenAI, but costs will not kill them. Text inference is a pretty safe bet.
The development of the last years has shown an exponential decrease in cost for inference at the same level of quality, partially due to smaller models of the same quality, partially due to optimizations in the pipelines. GPUs also get more efficient over time.
If you now sell a person a $20 subscription for text generation and keep the price for the next years, your costs will go down while the subscription keeps paying you
Open AI will likely succeed technically (Score:2)
They are doing excellent work and I suspect that the results will get better
Whether or not they ever achieve profitability is a different question
Investors are irrationally betting on achieving vast riches
It's entirely possible that AI will become very useful without generating the returns that investors expect
So, it's a financial bubble and solid, real technical progress
potential to disrupt the smartphone market (Score:2)
It already has - lots of people are already ditching their smartphones and I certainly won't be buying another now that Google wants to further restrict my sideloads.
No (Score:2)
The ignorant still whinge about 2008, as if it'd be fine for say, the US FDIC to bail them out and add to the national debt (which they'd pay for anyway) because they put their money in a shitty fucking bank, but heaven forbid the FDIC bail the bank itself out with a loan that got paid back (which they don't pay for).
OpenAI on the other hand can be instantly and seamlessly replaced by just going to any of the other chatbots that do the same damned thing. Do not pass Go, do not collect $200 bailout money.
Measuring failure? (Score:2)
Pretty weak FP. I think it is some kind of pre-loaded rant against fiat currency. Or maybe was an intended recursive joke about futures on futures? Insurance ^n as n approaches infinity? The Subject was certainly unhelpful. Maybe you care to clarify?
But I'm going to jump in a different direction: How do we tell if AI is failing. I think we are using the wrong metrics, so I would like to suggest a few candidates:
Best apologies: So far I think that one goes to Microsoft's Copilot for some stuff it said about