Is the AI bubble too big to burst?

Speaker 1 ABC Listen, podcasts, radio, news, music, and more.

Speaker 2 Hi, I'm Patricia Carvallis, the host of Politics Now, the ABC's podcast unpacking the latest political news when it happens. Every weekday, you'll hear me and the sharpest political minds at the ABC.

Speaker 6 Me, Jacob Greber, me, Raph Epstein, me, Claudia Long, me, David Spears, and me, Melissa Clark, in the chair for Fran Kelly for the party room, every every Thursday.

Speaker 2 So join me for Politics Now, download and follow on the ABC Listen app.

Speaker 1 This podcast was produced on the lands of the Owabacal and Gadigal people.

Speaker 1 About a year ago, two dudes in Silicon Valley launched a new podcast.

Speaker 8 Welcome to Technology Brothers, the most profitable podcast in the world. You read the Wall Street Journal today?

Speaker 9 Of course.

Speaker 1 They're wearing very casual clothes. They're sitting around a dining table.
They're extremely unscripted.

Speaker 9 Are they covering the departure?

Speaker 8 Not the first. Yeah, it's interesting because OpenAI,

Speaker 8 I mean,

Speaker 8 the chief technology officer quit yesterday.

Speaker 1 This episode has 2,000 views. Now, look, I've been around in the podcast game for a while now, and this doesn't strike me as a show that's going to rocket to the top of the charts.

Speaker 1 And yet, a year later, you're watching

Speaker 8 TBPN. Today is Wednesday, October 22nd, 2025.
We are live from the TBPN Ultra Dome, the Temple of Technology, the Fortress of Finance, the capital of capital.

Speaker 1 TBPN, as they call themselves nowadays, is Silicon Valley's favorite show. Several tech CEOs have actually made guest appearances.

Speaker 8 Mark Zuckerberg, live on TBPN. Sam Altman and Bill Peebles.
Sam, Bill, how are you doing? And we have Mark Cuban joining the stream.

Speaker 1 The view count on each episode is still quite low, like less than 10,000 views on most of their videos.

Speaker 1 But what makes this show so successful is who those viewers are silicon valley big dogs with a lot of money to spend so why are the big dogs so into this show well it's possibly partly because they let the weird tech bros come on and voice strange opinions like this guy who appears to be roasting himself i just started a new sauna protocol okay and it was 200 degrees fahrenheit 200 degrees and this guy kids should be having should be having kids when they're when they're in their teens

Speaker 11 That's when they're supposed to have them.

Speaker 1 Pretty creepy thing to say, my dude. But anyway, the most important thing about this podcast is that it's become the center of what they call the AI talent wars.

Speaker 8 The talent wars continue to rage on. Sam Altman slams meta AI's talent poaching spree mission.

Speaker 1 See, the AI industry is growing so fast that there's enormous demand for the software engineers and researchers who actually understand it.

Speaker 13 These are like the actual researchers who have been poached from various labs. So you see like Jason Wade, Lucas Feyer from OpenAI.

Speaker 1 TBPN is where all the trades, poaching and new signings are announced. And the salaries that some of these new hires are getting are truly insane.

Speaker 10 Apparently some of them are up to 300 million over four years with more than 100 million in total compensation for the first year.

Speaker 1 So TBPN is kind of sports center for AI.

Speaker 1 In July this year, Meta CEO Mark Zuckerberg signed a 24-year-old researcher named Matt Dietke for $250 million to head up a group called Meta Superintelligence Lab, and then apparently fired basically everyone else there.

Speaker 9 Meta cuts 600 jobs at AI Superintelligence Labs. The layoffs do not affect Meta's newest AI hires, who are in some cases being paid up to hundreds of millions of dollars.

Speaker 1 Now, I don't know, but all this feels a bit off. Vibes are weird, Chad.
And I don't mean from the guy baking himself in a sauna or the other guy advocating teen pregnancy.

Speaker 1 24-year-old software researchers are being treated like NBA players. And meanwhile, Meta, Facebook's capital expenditures are projected to surge to $72 billion in 2025.

Speaker 1 That's a $30 billion increase from the previous year.

Speaker 1 NVIDIA, the company that makes most of the microchips running the AI industry, has seen its stock rise 1,226%

Speaker 1 in three years.

Speaker 1 Two dudes are getting write-ups in the New York Times and Wall Street Journal for a podcast with 5,000 viewers. So what's going on here? How long can this absolutely massive growth continue?

Speaker 1 Well, if you watch shows on YouTube other than TBPN, you're going to hear a lot of people using the B-word.

Speaker 8 The entire AI industry is a bubble primed to burst. It seems like bubble mechanics to me, you guys.

Speaker 14 Whenever this bubble pops, there's going to be tens, if not hundreds of billions of dollars that will literally be incinerated.

Speaker 1 Literally? I think he means figurative. Never mind, doesn't matter.

Speaker 1 When people talk about AI being a bubble, they often compare it to the last big tech bubble known as the dot-com bubble, which burst rather dramatically 25 years ago.

Speaker 1 So today, what was the deal with that bubble? And does it tell us anything about what's going on with AI today? I'm Matt Bevan, and this is If You're Listening.

Speaker 1 So on the 6th of August 1991, computer scientist Tim Berners-Lee launched the first ever experimental website.

Speaker 1 For this, he's often called the inventor of the World Wide Web, although he's quite bashful about that title.

Speaker 15 Very lucky with the timing, very lucky with the way a lot of people picked it up and supported it and became a community which really developed it and brought it out.

Speaker 1 But given the size the internet is now, it's kind of surprising how slowly it grew in the early years.

Speaker 1 While there was a lot of hype, most companies were wary about trying to set up websites themselves.

Speaker 16 We didn't want to have an internet presence in-house. So we contracted with an outside service provider to help us to establish a World Wide Web site.

Speaker 1 This guy is from the Australian branch of the professional services company KPMG.

Speaker 16 I think it cost us about $5,000.

Speaker 16 We were very, very happy with the first couple of days. We had quite a few thousand hits, if you like, on

Speaker 16 our World Wide Web page, and that was a pleasant surprise. We had no idea how many people would access it.

Speaker 1 By the end of 1994, there were only 10,000 websites registered, which meant that huge numbers of common words were still available to you registered as domain names.

Speaker 1 Greg McLemore was in his mid-20s and owner of a company called Webmagic, which bought up potentially popular domain names like toys.com or pets.com and then sold them to companies once those companies figured out the internet was going to be a big thing.

Speaker 1 You've got to remember this was before search engines, so getting people to come to your website was tricky. A domain that was easy to remember was great for business.

Speaker 19 There's so much information out there on the web that being there isn't enough. Your electronic customers have to be able to find you.

Speaker 1 Having a domain name like toys.com or pets.com was potentially very valuable.

Speaker 6 If you haven't got a dot com, you'd better make plans to get one.

Speaker 1 The motivation behind those easy to remember domain names was similar to the motivation of businesses who build their stores on the side of the highway or right in the center of town.

Speaker 1 A prominently placed shop will usually outperform one that's on an obscure cul-de-sac on the wrong side of the tracks. And an obvious domain name would outperform more obscure ones as well.

Speaker 1 Or at least that was the theory. On the 20th of May 1999, Greg McLemore was about to test that theory.

Speaker 1 He had sold Toys.com to a company called e-toys in exchange for 750,000 shares.

Speaker 8 Where will you find the perfect gift for your child? Visit e-toys

Speaker 8 and it will come to you.

Speaker 1 And that day, e-toys was set to debut on the stock market. They offered shares at $20 each, which would have made McLemore an instant millionaire.

Speaker 1 But as it turned out, $20 a share was a massive lowball.

Speaker 1 Because by the end of the day, shares were trading at $76,

Speaker 1 making his stake worth around $57 million.

Speaker 1 So why were people buying up shares like Beanie Babies, Pokemon Cards, Pogs, and other 90s collectibles that would definitely increase in value?

Speaker 1 Well, the launch of e-toys came in the middle of a frenzy.

Speaker 4 In 18 months, almost any American company associated with the internet has soared to dizzying heights.

Speaker 1 The numbers were frankly ridiculous.

Speaker 4 Yahoo shares rose 1,000% last year and despite a recent dip are up 50% this year.

Speaker 4 Internet book retailer Amazon is worth more than all America's bookstores put together even though it hasn't made any profit.

Speaker 1 There was already significant scepticism about the amount of investment that was going into these companies.

Speaker 4 I don't know of anybody that can look at what's happened to internet stocks in the US and say, oh this is all reasonable, this all makes good, sound, fundamental economic sense.

Speaker 4 Bear in mind for most of these companies companies, you can't even calculate a so-called PE price-to-earnings ratio because most of these companies have no earnings.

Speaker 1 That doesn't sound like a winning mentality to me. As Christmas 1999 approached, the stock price for e-toys looked strong, but then they made a serious tactical error.

Speaker 1 Someone at e-toys noticed that a group of Swiss artists had a similar website name to theirs, eToy.com. This was extremely unfortunate for e-toys because these artists were, well...

Speaker 18 ETOY is a group of surreal business people.

Speaker 18 We are in the business of playing with corporate giants, with exciting playgrounds on the internet.

Speaker 1 They were extremely European.

Speaker 14 Money and shares are objects of our desire.

Speaker 1 Art is what we would kill for. E-Toys tried to buy the domain.

Speaker 6 ETOY Inc.

Speaker 3 offered the Zurich Activists 500,000 US for the name, which was crucial to its business plan.

Speaker 7 When the Zurich activists refused to accept the money, eToy Inc. threatened to sue and so the internet war began.

Speaker 1 12 days before Christmas 1999, the Swiss Art Collective launched a denial of service attack, basically getting thousands of people to access eToys.com all at once to jam up their servers.

Speaker 3 50,000 online activists jammed the dot-com's online shopping website, stopping genuine shoppers from buying toys before Christmas.

Speaker 1 It did not go well for e-toys.com. The costs of even minor disruptions to electronic commerce run into the billions of dollars.
By the time the toy war was over, their share price had dropped from $45

Speaker 1 to $6

Speaker 20 and it never recovered. E-Toy suffered tremendous financial difficulties after a disastrous Christmas season and has been laying off most of its workforce in recent weeks.

Speaker 20 Company insiders suspect the debt is close to $300 million.

Speaker 1 Still, that was just down to bad luck, right? They picked a name for their site that was very similar to a bunch of geopolitically neutral yodeling alpine jerks.

Speaker 1 Next time it'd surely go more smoothly. McLemore still had another cart up his sleeve.

Speaker 1 In early 2000, Pets.com was drumming up hype ahead of its stock market debut with ads featuring a sock puppet dog. I hope they're home.

Speaker 1 Oh wow, you cats hit the jackpot.

Speaker 1 There's enough food here to feed a lion.

Speaker 21 Pets.com because pets can't drive.

Speaker 1 The company seemed to have a lot going for it on paper.

Speaker 1 McLemore had hired an experienced tech boss named Julie Wainwright as CEO, and she had secured significant funding from Amazon boss Jeff Bezos in exchange for a 50% stake in the company.

Speaker 1 Their market analysis indicated that Pet Supplies in America was a $23 billion market, and they expected that through their website they would be able to carve off a big slice.

Speaker 1 That being said, early signs weren't great. In the three months leading up to their big stock market debut, they had brought in $5.6 million in revenue, that's good, but they'd spent $48 million.

Speaker 1 That's bad. They primarily spent that money on promoting the sock puppet dog.
But even without the sock puppet dog, they were losing money on every single sale.

Speaker 1 Competition in the pet supplies market was massive and the established retailers with physical stores like PetSmart were also setting up websites of their own.

Speaker 21 PetSmart.com uniting pets

Speaker 21 and delivery professionals everywhere.

Speaker 1 To undercut the competition pets.com was selling their products at significantly less than the price they cost to produce.

Speaker 1 The writing was already on the wall, even in the weeks leading up to its stock market debut. People were already starting to use that B-word.

Speaker 22 The rash of internet and technology companies floating at a handsome premium has many commentators worried that the investment bubble will soon burst.

Speaker 22 Analysts agree a shakeout of the sector is inevitable, but warn against tarring all internet and tech stocks with the same brush.

Speaker 1 Not all of the tech investment was going into dot-com businesses, though.

Speaker 1 Hundreds of millions of dollars was being spent by telecommunications companies building data centers and laying fiber optic cables which would allow the dot-com companies to operate.

Speaker 1 But despite a potential bubble burst looming, Pets.com pushed ahead and launched. On the 11th of February 2000, the stock market valued the company as being worth $290 million.

Speaker 1 At this point, tech stocks were making up more than a third of the value of the S ⁇ P 500 index, up from one-tenth just five years earlier.

Speaker 1 This enormous market share was a strong indication that something was up. It indicated that while most of the American economy was pretty stable, tech stocks were dangerously overvalued.

Speaker 1 In economic circles, alarm bells were starting to go off. The U.S.
Federal Reserve was concerned about what the chairman Alan Greenspan called irrational exuberance. Irrational exuberance.

Speaker 1 And in an effort to cool down the market, the Fed started raising interest rates.

Speaker 1 On the 13th of March that year, one month after the launch of Pets.com, the Yale University economist Robert Schiller published a book called Irrational Exuberance, and he was interviewed about it on the ABC.

Speaker 8 Is this a bubble market then? I mean, in these technology stocks, are we seeing? I don't think so.

Speaker 8 And if it is, does it mean automatically that one day it will burst?

Speaker 12 I don't think that bubbles burst typically in a catastrophic event. News media like to talk about one-day events like the crash of 29 or 87.

Speaker 12 What's more likely is a gradual decay of the level of prices and poor performance over years.

Speaker 1 Coincidentally, as he was saying that, literally that day, the bubble was bursting.

Speaker 22 The NASDAQ Index, the U.S. barometer which measures technology stocks, took its second largest drop overnight, falling more than 200 points.

Speaker 1 The era of irrational exuberance was over.

Speaker 17 Maybe some of them got to be a little bit overpriced or overvalued, and some of the stories that drove those Nasdaq tech-type stocks upward to some fantastic high valuations appear to be starting to be questioned.

Speaker 1 The gradual decay and poor performance over years didn't really eventuate. Tech stocks sank like a stone and none went down harder than Pets.com.

Speaker 5 The specter of higher inflation and rising interest rates led the frantic sell-off. The NASDAQ is now 34% down on its record high on March the 10th.

Speaker 1 Pets.com declared bankruptcy in November 2000, nine months after it debuted on the stock market. $300 million in investors' money had evaporated.

Speaker 1 The following year, the companies that had gone into massive debt building data centers and the fiber optic cables to support companies like Pets.com followed them into the abyss.

Speaker 6 The news that Telecom's giant WorldCom was close to going bust had investors running for the hills.

Speaker 1 So why did this happen? Well, in part, it was because of the kind of group hysteria that leads to every stock market bubble, the irrational exuberance.

Speaker 1 But it was also a misunderstanding about the value of a memorable domain name.

Speaker 1 Search engines may not have existed when Greg McLemore bought those domain names in 1994, but they definitely existed when he actually launched the companies five years later.

Speaker 22 Yahoo isn't just a Homer Simpsons saying, it's also a search mechanism that helps you find what you want on the World Wide Web.

Speaker 1 And even before the bubble got to stratospheric size, the vast majority of people were finding websites through search engines and links from other websites.

Speaker 1 Neither of those methods value a good domain name. They value a good website and a good user experience.

Speaker 1 PetSmart, the big, well-established pet supplies company, was able to set up a quality site and provide quality service. So why would anyone need to go to pets.com?

Speaker 1 In the end, pets.com was bought for pennies on the dollar by PetSmart. eToys.com was bought for a similar amount by Toys R Us, the big, well-established retailer in that industry.

Speaker 21 The days of buying anything with a dot-com are over.

Speaker 1 While most of the companies impacted by the dot-com burst vanished into obscurity, a few of the bigger companies managed to weather the storm, like Amazon.com with its CEO, Jeff Bezos.

Speaker 24 I think really what you'll see is there start to be a division between companies that have business plans that really fundamentally make sense, and those companies can create lasting companies and investors will continue to reward those companies.

Speaker 24 And then a second set of companies that were trying to capitalize on sort of, you know, internet fever.

Speaker 1 As we all know, Amazon and Jeff Bezos survived and eventually figured out a way to turn a profit. But Bezos has not forgotten that era because it was pretty traumatic for him.

Speaker 25 In the year 2000, when the internet bubble burst, Amazon's stock in a very short period of time went from $113

Speaker 25 a share to $6 a share.

Speaker 1 And he can see certain similarities with what's going on now.

Speaker 25 Every experiment gets funded.

Speaker 25 Every company gets funded. The good ideas and the bad ideas.

Speaker 1 A huge amount of money is being thrown at anything with AI in the name, just like in 1999 with anything with dot-com in the name.

Speaker 25 Investors have a hard time in the middle of this excitement distinguishing between the good ideas and the bad ideas. And so that's also probably happening today.

Speaker 1 In July 2025, tech companies' share of the S ⁇ P 500 exceeded one-third for the first time since the height of the dot-com bubble. So that's kind of ominous.

Speaker 1 But Bezos says that this time around with AI, things are different because there's something tangible at the center of the hysterical speculative investment.

Speaker 25 Like, AI is real, and it is going to change every industry. The biggest impact that AI is going to have is it is going to affect every company in the world.

Speaker 25 It is going to make their quality go up and their productivity go up.

Speaker 25 By every company, I literally mean every company, every manufacturing company, every hotel, every, you know, consumer products company, etc., etc., etc.

Speaker 25 And so

Speaker 1 That is hard to fathom, but it's real.

Speaker 1 It's real.

Speaker 1 And by that he means that while Pets.com may have sunk most of its money into marketing a sock puppet dog, AI companies are sinking at least some of their money into physical infrastructure.

Speaker 1 He says the current race to pour money into AI innovation is less like investing in individual sites like Pets.com and more like investing in the telecommunications companies that supported them.

Speaker 25 All of that fiber optic cable that got laid, and by the way, the companies who laid all that cable went out of business. business.

Speaker 8 Like literally went bankrupt.

Speaker 25 But the fiber optic cable was still there and we got to use it.

Speaker 1 The obvious question is though, will the infrastructure being built during this AI boom, the massive data centers, high-speed networks and processing facilities, be as useful as those fiber optic cables if and when some of these companies go bust?

Speaker 1 Is AI actually something that someone will be able to make money from in the long term?

Speaker 1 Because despite the huge investment the big tech firms are making into new young staff and all these data centers, they sure aren't making any money from it now.

Speaker 1 And even if it does make money, is a world with widely available AI sustainable?

Speaker 1 If the AI systems are as hungry for electricity and water and microchips as everyone says, will they and should they be used in the future? We'll look into that next on if you're Listening.

Speaker 1 If You're Listening is written by me, Matt Bevan. It's produced by Adair Shepherd.
Supervising producer is Cara Jensen-McKinnon.

Speaker 1 On Tuesday, Cara has been sucked into the world of the Pets.com sock puppet dog. In the midst of the implosion of the company, the dog got into a literal legal battle against a different dog puppet.

Speaker 1 And the legal battle got kind of heated.

Speaker 1 In fact, the dot-com bubble spawned a whole lot of very strange lawsuits. It's a bizarre story, and that's what we'll look into on Tuesday.
I'll catch you then.