There's No 'AI Bubble', Says Yahoo Finance Executive Editor (yahoo.com) 68
"I'm here to say we have to give these AI bubble predictions a rest," says Yahoo Finance executive editor Brian Sozzi.
First of all, AI is a real technology being deployed in real ways inside of Corporate America. Second, this technology is requiring more physical assets in the ground — which are being built to support AI's real-world application. What Zach Dell (son of Michael Dell) is working on at startup Base Power (which just raised $1 billion) impressed me this week. It's addressing a key issue — power availability and costs in part because of rising stress on the grid due to AI development.
Next, the spending on AI infrastructure doesn't strike me as reckless. I talk to CFOs and they walk me through their thinking, which seems logical. They aren't foaming at the mouth with wild-eyed predictions of grandeur similar to the late '90s. Plus, the tech giants making the biggest AI investments are fueling their ambitions by cash on hand — not loading up balance sheets with debt. The upstarts in AI are well funded, not being 100% stupid in their organizational build-outs. They're working on tangible technology that has actual orders behind it...
Lastly here in my scolding of the AI worrywarts is that valuations don't support the warning calls. According to new research out of Goldman Sachs this week, the median forward P/E ratio across the Magnificent Seven is 27 times, or 26 times if excluding Tesla (TSLA), which has a much higher multiple than the other companies. This is roughly half the equivalent valuation of the biggest seven companies in the late 1990s, while the dominant companies in Japan (mostly banks) traded at higher valuations still. What's more, the current enterprise-to-sales ratios are also much lower than those of the dominant companies in the late 1990s.
"So it is true that valuations are high but, in our view, generally not at levels that are as high as are typically seen at the height of a financial bubble," said Goldman Sachs strategist Peter Oppenheimer.
Next, the spending on AI infrastructure doesn't strike me as reckless. I talk to CFOs and they walk me through their thinking, which seems logical. They aren't foaming at the mouth with wild-eyed predictions of grandeur similar to the late '90s. Plus, the tech giants making the biggest AI investments are fueling their ambitions by cash on hand — not loading up balance sheets with debt. The upstarts in AI are well funded, not being 100% stupid in their organizational build-outs. They're working on tangible technology that has actual orders behind it...
Lastly here in my scolding of the AI worrywarts is that valuations don't support the warning calls. According to new research out of Goldman Sachs this week, the median forward P/E ratio across the Magnificent Seven is 27 times, or 26 times if excluding Tesla (TSLA), which has a much higher multiple than the other companies. This is roughly half the equivalent valuation of the biggest seven companies in the late 1990s, while the dominant companies in Japan (mostly banks) traded at higher valuations still. What's more, the current enterprise-to-sales ratios are also much lower than those of the dominant companies in the late 1990s.
"So it is true that valuations are high but, in our view, generally not at levels that are as high as are typically seen at the height of a financial bubble," said Goldman Sachs strategist Peter Oppenheimer.
Consider the Source (Score:5, Funny)
After all, who would know more than Yahoo about over-inflated valuation [qz.com]?
Re: (Score:3, Informative)
The real tragi/comedy is that we still have Goldman Sachs on a pedestal. They are the one of the banks that caused the last big bust. They have nothing to worry about. They'll just get bailed out, again
Bunch of pump 'n dump scammers
Re:Consider the Source (Score:5, Insightful)
Wasn't Goldman Sachs one of the investment banks that started to short the mortgage backed security business for its internal accounts while still pushing them on customers?
Re: (Score:2)
Re:Consider the Source (Score:5, Informative)
By Pam Martens and Russ Martens: December 15, 2016
On Wednesday, April 13, 2011, following a two-year investigation, Senators Carl Levin and Tom Coburn, Chairman and Ranking Member of the Senate’s Permanent Subcommittee on Investigations, released a 635-page report which included specifics on the deceitful and fraudulent role that Goldman Sachs played, among others, in burning down Wall Street and the U.S. economy in the greatest collapse since the Great Depression.
Goldman Sachs is referenced 2,495 times in the report; Cohn is referenced 89 times. Cohn’s role in overseeing the firm’s peddling of its subprime mortgage-backed securities is called out in the example below; (other references show how Cohn kept close tabs on the firm’s shorting of the market):
“In 2006 and 2007, the head of the Mortgage Department was Daniel Sparks. Goldman Co-Presidents Gary Cohn and Jon Winkelried, as well as CFO David Viniar, had been involved in Mr. Sparks’ earlier career at Goldman, and he maintained frequent, direct contact with them regarding the Mortgage Department.”
The report characterizes the actions of the denizens of Wall Street as an “economic assault” on the United States. The report came a year after the Subcommittee had conducted a series of hearings in April 2010, including taking direct testimony from Goldman Sachs officials. (See video clip below.) The report summarizes its findings about Goldman Sachs as follows:
“When Goldman Sachs realized the mortgage market was in decline, it took actions to profit from that decline at the expense of its clients. New documents detail how, in 2007, Goldman’s Structured Products Group twice amassed and profited from large net short positions in mortgage related securities. At the same time the firm was betting against the mortgage market as a whole, Goldman assembled and aggressively marketed to its clients poor quality CDOs that it actively bet against by taking large short positions in those transactions. New documents and information detail how Goldman recommended four CDOs, Hudson, Anderson, Timberwolf, and Abacus, to its clients without fully disclosing key information about those products, Goldman’s own market views, or its adverse economic interests. For example, in Hudson, Goldman told investors that its interests were ‘aligned’ with theirs when, in fact, Goldman held 100% of the short side of the CDO and had adverse interests to the investors, and described Hudson’s assets were ‘sourced from the Street,’ when in fact, Goldman had selected and priced the assets without any third party involvement. New documents also reveal that, at one point in May 2007, Goldman Sachs unsuccessfully tried to execute a ‘short squeeze’ in the mortgage market so that Goldman could scoop up short positions at artificially depressed prices and profit as the mortgage market declined.”
The report does a deep-dive into how Goldman Sachs actually created product for the express purpose of reducing its own exposure to subprime mortgages while offloading the toxic paper onto its clients. The report notes:
“The review begins in mid to late 2006, when Goldman realized that the market for subprime mortgage backed securities was beginning to decline, and the large long positions it held in ABX assets, loans, RMBS and CDO securities, and other mortgage related assets began to pose a disproportionate risk to both the Mortgage Department and the firm. In October 2006, the Mortgage Department designed a synthetic CDO called Hudson Mezzanine 2006-1, which included over $1.2 billion of long positions on CDS contracts to offset risk associated with ABX assets in Goldman’s own inventory and another $800 million in si
Re: (Score:2)
Probably should’ve bought them. [youtube.com]
Re:Consider the Source (Score:5, Insightful)
What's your source for that?
The people who are creating the bubble.
Oh, OK. Then I'm sure everything will be fine.
Idiots.
Re: (Score:3)
"AI is really used for actual real problem solving like mediating the power grid stress problem caused by AI usage"
That's like the recursive mafia version of the broken window parable...
Then it goes on with being convinced by seemingly logical arguments that are not 100% stupid (what a relief) and concludes with some nice whataboutism, others do it too, Tesla is even worse !
Just wow...
Professional liar says what? (Score:5, Funny)
Re: (Score:2)
Hey Mr, Altman, nice to see you posting on Slashdot. I find it really strange, however, that you can't seem to type capital letters. I mean, you are clearly too well-educated to misunderstand their use. And certainly wealthy enough to afford a working keyboard or phone. Is it some idiosyncrasy on your part? Like, someone who is customarily as busy as you just doesn't have the time to waste on things like holding shift or tapping capslock as you type along?
In any event, I am working on some LLM-powered
Re: (Score:2)
I'll wait for Sam Altman to reassure me that there is no bubble.
Ironically, Altman agrees that AI is a bubble [cnbc.com]:
OpenAI CEO Sam Altman thinks the artificial intelligence market is in a bubble, according to a report [theverge.com] from The Verge published Friday.
“When bubbles happen, smart people get overexcited about a kernel of truth,” Altman told a small group of reporters last week.
“Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes. Is AI the most important thing to happen in a very long time? My opinion is also yes,” he was quoted as saying.
Altman appeared to compare this dynamic to the infamous dot-com bubble, a stock market crash centered on internet-based companies that led to massive investor enthusiasm during the late 1990s. Between March 2000 and October 2002, the Nasdaq lost nearly 80% of its value after many of these companies failed to generate revenue or profits.
Nobody is arguing it's not a real tech (Score:3, Insightful)
It's not a traditional bubble where there isn't anything there it's a bubble because markets consolidate to crazy levels nowadays leaving only a couple of players and everyone else caught holding the bag
Re: Nobody is arguing it's not a real tech (Score:5, Insightful)
Even the housing bubble had real houses in it. I don't get how people at this level can be so dense or think the rest of us are so stupid.
Re: Nobody is arguing it's not a real tech (Score:5, Insightful)
This. The dot-com bubble was still a bubble, but there were real companies producing real websites, some of which were even useful.
It's the same thing with AI. The fact that the technology exists and sometimes is used for things that are useful doesn't change the fact that there's a *huge* hype bubble around AI, and everybody and their mother is dumping piles of money into AI, hoping that they'll get lucky and back one of the winning horses. All the while, companies with no real business plan other than "AI" or "AI first" or whatever are hoping the investors won't notice that the companies really don't know what to do with AI, and they're just hoping that if they build it better, first, they'll "win" or whatever.
It's a bubble. It's a huge bubble. I have no idea when it will burst, but it will. They always do.
Re: (Score:2)
Most of the people on the internet at the time remember Paypal. For a while, you couldn't buy or sell on ebay without it (unless you wanted to wait for a money order or paper check to clear). Is there anyone that remembers X.com as an online bank? I sure don't. But somehow this site https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fweb.archive.org%2Fweb%2F20... [archive.org] got Peter Thield to buy Elon Musk's X.com before Paypal was sold to Ebay. Apparently it was only a "real
Re: (Score:2)
This. The dot-com bubble was still a bubble, but there were real companies producing real websites, some of which were even useful.
Indeed! And not only that it was real tech, proved to be immensely useful, changed the world and ultimately saw absolutely colossal growth. But as you say, still a bubble.
It's not the first time, either. One could say the same about powered transport, too. In 1829, the first powered wheeled vehicle chuffed along with passengers. No one looking back from now could argue that this
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
There are two players only.
Open AI and Nvidia, everyone else is giving money to them.
Translation (Score:5, Funny)
I mostly left the stock market on Friday (Score:2)
How are the prospects for an early earthquake in California... ;)
Re: I mostly left the stock market on Friday (Score:2)
About the same as a stock market crash of the same dignity as the '29 crash.
What silly arguments... oh, yeah, it's Yahoo (Score:4, Insightful)
There was lots of real technology being deployed prior to the burst of the DotCom Bubble.
There were a lot of people who thought the money they were dumping into companies like Kozmo.com was not being spent recklessly.
What's funny here is his basic argument devolved into "okay, yeah, the valuations are higher than they should be... but I think it's okay!"
Re: (Score:2)
P.S. I do still miss Kozmo.com - it was great while it lasted!
Re:What silly arguments... oh, yeah, it's Yahoo (Score:5, Interesting)
The dotcom bubble was more of a "internet shiny thing"
All these companies who were not yet on the internet, were getting on the internet. A lot of companies that should never have existed, also got on the internet.
The "the cloud" bubble was another one, with all these moronic companies outsourcing to data centers so they don't need to own the hardware, and then have their ass handed to them when someone forgets to pay a bill, or the data center has poor security.
The AI bubble is far bigger and incestuous between companies producing chips (Nvidia, AMD, and Intel) and companies deploying "AI" in data centers.
Like we've come around full circle on the the reasons why it's a giant ponzi scheme of dumping money into something and the investors are going to eventually fall on a bed of swords.
Re: (Score:2)
IN '21-'23 the NBT was the "Metaverse"...
To the extent that Facebook re-named itself "Meta" and they even released commercials on how the Oculus was the future of office work.
"Sure, we're making all our employees be physically sitting at a cubicle, in an office building, but these shitty goggles are the future of office work!"
Meta thus spent over 100 billion dollars on the "metaverse". a year later, after the release of ChatGPT, Meta threw the "metaverse" in the garbage and now the future of everything is L
They are getting truly desperate now (Score:5, Insightful)
Direct lies is usually the last means to keep the hype going and a sign of imminent collapse.
Re: (Score:2)
Re: They are getting truly desperate now (Score:4, Insightful)
Well, if your customers (in this case: voters) are really, really, really stupid, this can work a bit longer.
Re: (Score:2)
The common man, no matter how sharp and tough, actually enjoys having the wool pulled over his eyes, and makes it easier for the puller.
P. T. Barnum
Re: (Score:2)
Indeed. Sad as that is, because it means democracy is easy to corrupt because people are fucking dumb.
Re: (Score:2)
Probably no coincidence he felt compelled to write after the Friday crash. Wasn't related to AI but since about 35% of the S&P 500 is concentrated in AI concerns, it's particularly vulnerable to general stock disruption.
It's two things (Score:4, Interesting)
Real, rapid progress is being made toward creating useful tools that will help scientists and engineers solve previously intractable problems.
There is an irrational frenzy, driven by hypemongers and pundits that cause investors and gamblers to pour billions into anything with AI in the name as they desperately look for "the next big thing".
So yeah, it has a strong bubblish aroma
Re: (Score:2)
Real, rapid progress is being made toward creating useful tools that will help scientists and engineers solve previously intractable problems.
Yes, we can call this actual artificial intelligence. Good work is going on, but it's not the stuff being hyped.
There is an irrational frenzy, driven by hypemongers and pundits that cause investors and gamblers to pour billions into anything with AI in the name as they desperately look for "the next big thing".
Yes, this is the hype: large language models and their cousins, which have no actual intelligence, but simply put together patterns.
So yeah, it has a strong bubblish aroma
Re: (Score:2)
Monkeys like to talk to each other. LLMs do that, so they get press coverage. They're not useless either: we have an enormous amount of natural language data that was previously very difficult to deal with and is now much more accessible to conventional computing, from the freeform comment boxes on surveys to physician notes and regulatory documents and scientific literature.
Do not confuse what you read in the popular press, or what Slashdot editors post as clickbait, with what's actually being done.
Doesn't know what a bubble means? (Score:5, Insightful)
The internet was also real, that doesn't mean we didn't have a dot-com bubble.
Re: (Score:2)
I'm also pretty sure houses were real.
WEAK (Score:1)
Just saying "the valuations are fine", "the tech is useful", "the infrastructure is needed" is so lame a defense, he's basically telling us it's farked.
Re: WEAK (Score:2)
Nothing to see here, keep investing! (Score:4, Insightful)
Makes me wonder which stocks this guy is investing in.
Re: (Score:2)
NVDA Bubble (Score:5, Interesting)
there is an NVDA bubble, as the GPUs are being sold at outrageous premiums. It's bubbling left and right with de-facto vendor financing (2000 deja-vue) though. Which is not sustainable since a) competition will inevitably kick in, especially for the enormous profit potential as of today b) only a fraction of GPU functionality is needed for the linear algebra operations for LLMs c) hardware optimized for transformers including KV cache are/will be coming and not only from Nvidia, d) hyperscaler customers are showing reluctance to pay those premiums despite the CUDA bastion, at last going to other sources (AMD as of this week) besides in-house silicon.
Re: (Score:2)
At this point, about 35% of the SP500 is pretty directly predicated on AI. That's way too much across the board.
Moot point (Score:4, Insightful)
If corporate America has it's way, there won't be any pesky "workers" to worry about soon.
Sigh...
if you can't see the AI bubble (Score:2)
Then you're sitting on it.
Yahoo said there is no bubble? (Score:2)
Wow, this so ironic on so many levels!
Who else is old enough to remember Yahoo during the Dotcom bubble?
Re: (Score:2)
In September 1995, Cuban and Wagner then worked out a deal whereby Jaeb kept 10% of the company and got a monthly salary of $2,500 for part-time work but Cuban took control of the company, which was renamed AudioNet.com.[3] At first, Cuban picked up signals from KLIF (AM) in his bedroom and then streamed them on the Internet.
On April 1, 1999, less than nine months after the IPO, Yahoo! announced the acquisition of Broadcast.com, which then had 570,000 users.[9] The acquisition closed in July 1999 with Yahoo issuing 28.6 million shares of stock worth $5 billion for the Broadcast.com shares.[10] Cuban sold most of his Yahoo! stock that same year, netting over $1 billion
Yahoo! shut its broadcast services, including Broadcast.com, in 2002.[12] Yahoo!'s purchase of Broadcast.com has since been called one of the worst Internet acquisitions of all time
I don't understand ... (Score:2)
Is Sozzi trying to get me to buy AI services? Or invest in AI infrastructure?
That's not what bubble means... (Score:5, Informative)
First of all, AI is a real technology being deployed in real ways inside of Corporate America.
A bubble doesn't mean the technology is fake, it means it's being overvalued.
The Internet was (and is) a real technology being deployed in real ways inside of Corporate America. That didn't make the dot-com bubble not a bubble.
Second, this technology is requiring more physical assets in the ground — which are being built to support AI's real-world application.
So did the Internet; there was a telecommunications bubble at the time. This is not a good argument.
Re: (Score:2)
Most bubbles are rooted in something real, the dotcom, the housing market. You do have things like crypto that is more purely fictional, but most mainstream bubbles are rooted in real stuff..
Re: (Score:2)
Does it? I wouldn't be surprised if you looked at an index of dot com companies twenty years later and it came out with a pretty good return, even comparing to the height of the bubble. Certainly the "tech" industry, which is mostly dot com web software companies plus a few picks and shovels suppliers like Microsoft, has beaten the pants off other sectors.
Bubbles are a lot of people looking for a quick buck investing in a new thing
We need more nuclear energy (Score:1)
If the "AI bubble" is driving up the demand for electricity then we will need nuclear power plants built. I can anticipate a number of counter arguments already.
"What about the waste?"
You are being lied to about the problem of radioactive waste: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3F... [youtube.com]
That video I linked to may not be the best source but it is at least recent that it should avoid any complaints of being outdated. It's apparently a trustworthy enough source, but I don't know all that much of Business Insider to
Re: (Score:2)
18 Months for a "nuke", that is overly optimistic. I expect at least double that amount of time to let all the concrete dry out, let alone form it how it is needed to be. And no, most of that cannot be done off-site in several parts, all built at the same time. As that would be a logistical problem in the general area. And another big problem (transportation) of the elements that can be manufactured in other provinces/states.
Hinkley 2, that is a huge "nuke" in Great Britain, as replacement for Hinkley, the
Promise to wear a chicken suit if bubble (Score:1)
They are using their increased hype-supported stock price.
Let's see the actual orders. If you give it away for free (to gain market share) then demand will indeed appear, but what about when the subsidies dry up?
key word:Yahoo! (Score:2)
They said the same thing (Score:2)
Re: (Score:3)
"They" do. The Great Recession debt bubble was immediately proceeded by luminaries such a Barney Frank self-assuredly denying any problems [youtube.com], and characterizing the pesky noticers as misguided.
The AI bubble is end stage. LLMs are essentially next level search engines and, while powerful, their power is finite, especially since the training material is already exhausted. Further, there is little value in redundant implementations: there is no need for a dozen plus distinct tier 1 LLMs all exhibiting appro
Re: (Score:2)
Unfortunately looks like they are sufficiently impatient they've pivoted to trying to fire up more coal instead...
We from Yahoo recommend Yahoo! (Score:2)
Yahoo managers are just inflating the bubble a bit more in return for a nice quarterly bonus.
Say yahoo finance... (Score:2)
CFOs use it to cut labor... (Score:2)
so.... (Score:2)
He's saying we're in a bubble?
thought so.