Will the AI race fuel another boom or another bubble? | Counting the Cost

追加: 2026-05-31

[00:00:14]Hello, I'm Rashad Salam. This is Counting the Cost in Alazer, your look at the world of business and economics.

[00:00:21]A new front in the AI race. Some of the world's biggest tech companies are preparing to go public. But is this the start of a market bubble or the beginning of a technological revolution?

[00:00:31]A handful of companies are dominating AI. Investors are pouring billions into the technology. But is this the next great economic transformation or a disruption that leaves most people behind?

[00:00:47]The AI boom is entering a new phase.

[00:00:49]With OpenAI, SpaceX, and Anthropica are edging closer to becoming publicly traded companies in what could be one of the biggest waves of tech listings in history. Combined valuations already run into the trillions. Through pension funds, retirement savings, and managed investments, people could end up owning these shares in these companies, whether they choose to or not. These firms need a lot of money to build data centers, servers, and infrastructure. Going public is one way to raise it, but none of them have yet generated an annual profit and analysts warned their business models remain well opaque. So our people being asked to bankroll Silicon Valley's biggest gamble or biggest success. Finton Monahan reports.

[00:01:32]The management of AI chip maker Sarah Brass celebrate a blockbuster IPO.

[00:01:37]Stocks surged nearly 70% on their debut, putting the company's value at $67 billion.

[00:01:44]It was the biggest tech debut since 2019, and it could be just the beginning.

[00:01:50]Three more big players are expected to hit the stock market soon. SpaceX, Anthropic, and Open AAI. These are expected to hit a combined market capitalization of more than $3 trillion.

[00:02:02]Massive amounts of money has been pouring into AI and not just on the stock market. Tech giants like Google, Meta, and Microsoft are pouring billions into building data centers. High-end chipmakers like Nvidia are selling hardware faster than they can churn it out, and businesses are scrambling to integrate the new technology into their workflows. A report by Goldman Sachs estimated 300 million jobs could be affected by AI automation. It predicted a boost to the global economy of 7% over 10 years. The World Economic Forum put job losses by 2030 at more than 90 million, but predicted that more than 170 million new jobs would be created.

[00:02:43]But there's concerns that the frenzy of AI investment could be getting ahead of itself. The head of Bank of America is among those warning that a bubble may be forming. The market is becoming narrowly concentrated around the tech sector with tech firms on course to account for nearly half the value of the S&P 500.

[00:03:01]This happened before in the 1990s during the dotcom bubble.

[00:03:05]Excitement about the transformative potential of the internet overheated the market. This eventually led to a crash in the early 2000s. Tech stocks fell sharply and there were mass layoffs across the sector. Just like the internet, the AI revolution is likely to have a lasting impact on the global economy. But there's growing concern that if the market gets too far ahead of reality, it may lay the groundwork for a crash. Vinton Mahan, Al Jazer for counting the cost.

[00:03:34]>> So let's take a closer look at how companies are using the money being invested. The tech giants are plowing money into what they say is critical AI infrastructure. A combined $700 billion of spending is expected for 2026. AI firms themselves are spending big on expanding their operations. Open AAI are planning on shelling out some $50 billion this year. Private equities also getting in on the action. AI firms have raised $297 billion in the first three months of this year according to data from Crunchb. And most of it is concentrated in those very same companies mentioned earlier that are gearing up for IPOs. As we mentioned before, OpenAI, Anthropic, and SpaceX.

[00:04:14]Now, today's rapid rise in AI investments is raising questions about whether current valuations are justified or whether markets are heading toward another correction, and if so, how deep.

[00:04:24]What can history really tell us about moments like this? Now, in the run-up to the 1929 Wall Street crash, investors borrowed money under weak regulations and poured it all into booming stocks as speculation spiraled. By 1932, more than $30 billion, as 90% of the market's value had been wiped out. A similar rush surrounded internet companies before the dotcom crash in 2000 to 2002 that wiped off, would you believe it, $5 trillion in market value. Investors were putting money into businesses with no clear path to profit, betting future growth would justify soaring valuations. Now, before the 2008 financial crisis, cheap credit and risky lending fueled a housing boom that eventually collapsed. More than $10 trillion was wiped from global markets.

[00:05:11]Well, let's discuss all this further with our panelists. From New York is Daniel Ives. He is the global head of technology research at Wedbush Securities. From Los Angeles, Romesh Shini Vasan, professor at the UCLA Department of Information Studies and a specialist on AI and technology. And from Geneva, Adrien Monk, senior adviser on AI and tech to the United Nations and editor of the Seven Things Newsletter.

[00:05:35]Well, welcome to the program. Um, let's start off, Adrian, with you. What what are you making of these companies which are basically the software side of AI as opposed to the hardware which is of course these chip companies? Are we in bubble territory? That that's actually going to be the fundamental question here. Adrian, >> I mean, look, a 100red years ago, the great economist John Maynard Kanes before the 1929 crash said markets can stay irrational longer than you can stay solvent. And I suppose today's version of that is AI can stay unprofitable, you know, longer than investors can stay patient. And I think this is the fundamental issue behind this. There's a lot of money going in. Is that money going to result in productivity gains that we're going to see and be able to feel and that investors can be confident in? And at the moment, I think the jury is very much out on that. We don't see a lot of the evidence that these companies are going to pay back along the timelines that they're borrowing against. So, there's other risks too in this whole process. You know, public markets are designed for accountability.

[00:06:52]A lot of the accountability has been drained out of these processes. You won't get a vote if you buy into Elon Musk's SpaceX. You didn't get a chance to see if he wanted to lash X AI into it. So, some of the accountability in these public markets is missing from these IPOs and also the actual schedules under which we'll see the gains from AI are very very unclear right now. You know, Nvidia is selling chips onto Microsoft. Microsoft's giving that compute to open AI. Open AAI is then saying that's making Microsoft Azure better. You know, a lot of circularity here that rings a lot of bells for people old enough to remember 1999 when a lot of telecom's companies were doing similar sorts of things. So, it's very much uh a kind of wait and see for anyone looking to put money into this.

[00:07:43]Uh >> Daniel, your thoughts? I mean, would you be a buyer? I mean, on the other hand, would you prefer to be in the uh the picks and shovel side of things here as well rather than at the front end of the software side of things?

[00:07:58]Look, I mean, I get the points Adrian's making, but to me, I mean, this is a true fourth industrial revolution. And it's been our view the last few years, which is why this is going to be a multi-year tech driven bull market because for every dollar spent on Nvidia chip, there's an $8 to$10 multiplier across the rest of tech. And when you look at SpaceX, you look at Open AI, you look at anthropic, I mean, these are some of the foundational pieces to to what's really a true fourth industrial revolution. Now, look, what Adrian's bringing up, and I think what what many bring up, you got to see monetization.

[00:08:32]You got to see the use cases build out on the enterprise and on the consumer.

[00:08:37]That's right now going to be the execution phase. But I mean, I covered tech stocks in the late 90s. This is not a bubble. I continue to view it. This is probably, you know, call it 10 15% of the way through this AI revolution build out. Get the popcorn out.

[00:08:53]>> Romesh, your thoughts.

[00:08:56]>> I mean, no question that from an investor perspective, there are big questions about um what revenues and returns are going to come back uh to investors on what schedule uh related to these companies. Uh but it's astonishing as the report said, you know, we're seeing estimates of 50 to 60% of the entire S&P 500 being driven by these AI stocks. And I think it's good to step back and take, you know, take a look at the big picture here. So what we're seeing in my opinion with the investments here and the in the investor and private equity money going into this are bets, right? These are bets on which of these companies, especially in the cases of some of the softwarebased companies, are going to turn into the essential brains of AI systems that are going to be integrated into everything and everywhere. But I think the big question here is, you know, I mean almost a little more existential. Uh I was reading statistics that the world is more economically unequal than during the Roman Empire and certainly during feudal lord times. And I think that that's a big question, right? We have an astonishing amazing potentially incredibly uh productive technology for all of us that we have created that is using all of our data but no one's being paid for it and the economic insecurities that most are facing around the world and definitely the anxieties as we see here in the US with the incredible backlash around these data centers that are just not creating jobs.

[00:10:20]We're not seeing a lot of future planning with jobs and employment. We're not seeing a lot of in, you know, security here. And that reminds me of, you know, a time when John Maynard Kanes himself predicted that automated technologies would create a 15-hour working week. But we're not seeing that.

[00:10:37]We're seeing the vast majority of people here in the US and around the world working harder in a sense for less, and the money is just going to some of us who are investing otherwise.

[00:10:46]>> I want to delve into that a bit later.

[00:10:47]Dan, back to you for a second. I mean, you look at the S&P 500, we look at, you know, trailing price to earnings ratios of what, mid20s, if you will. Now, that's historically a little bit pricey, but not that pricey. You can't say it's that overvalued. And if you look at the socks, the uh semiconductor index, that's trading at similar sort of ratios as well. So, you know, there's that shows there not over exuberance, doesn't it?

[00:11:12]That's why investors continue to buy the dips because these are stocks that are still I think you know relatively cheap especially when you look at Avidia and others to to what the future's going to be over the coming years as the execution story plays out. Hey, look, I mean like Romesh, Asian bring up great points in terms of jobs. I almost say tech has a PR problem, you know, in terms of a lot of issues being created in terms of almost the AI alarmism. But the reality is like there's more data centers in the US under construction than active data centers. And what that means is whether it's Micron, whether it's Sandis, the infrastructure, the hyperscalers, the second, third, fourth derivatives, ultimately power and energy, I mean, we are still early in where this is going to play and stocks are not egregious and that continues to be our thesis.

[00:12:04]>> Well, Adrian, I mean, is the market pricing in the potential but not looking as ever and ignoring the risks? And what are those risks? Look, I think what we're seeing is a timing issue. You know, it's undoubtedly the case that this technology is going to be transformative and it's going to impact whole industries, jobs, sectors, you name it. The issue really is in the financing of that and how we actually pay for it and is that uh, you know, hoped for gain going to come on time and if it doesn't, that's going to affect a lot of the players that we're currently seeing in the marketplace. And I think you know go back in time and look at some of the companies that missed the boat. Uh you know the globe.com one of my former colleagues was uh was on the staff of that company which was like the biggest rocket ship of the 99 bubble and you know it missed. Now a lot of other companies if you'd had money in Amazon and people like that you'd have ended up doing very well over time but we don't really know who the ultimate winners and losers in this revolution are going to be. Right now people are saying it's going to be anthropic. It's going to be open AI and with Starlink and stuff uh slightly to one side it's going to be people like Elon Musk. Is that guaranteed? We don't know. There's a lot of geopolitical risk we've seen from the Hormuz closure. There's a lot of unpredictability. These are all the kinds of things that derail the bestlaid plans of investors and it would be very foolish to think that they won't be coming into play in this particular industrial revolution.

[00:13:36]>> Yeah. globe would one let's think go back in history and look at MySpace I mean whatever happened there so Romesh give us a sense of what you um think we can learn from the way the dotcom bubble burst and are those lessons applicable to AI >> I think some of the more recent lessons tied to big tech companies and social media companies are also really relevant here um you know as you alluded to um bets on Amazon ended up working out, right? But at the m but in the meantime for the a large part of its history and a company like Uber, these were companies that were uh revenue negative uh investor subsidized and certainly not profitable. Um but at the same time I think the question is which of these AI companies is really going to win in the sense of having the most refined model that is the most interoperable and applicable to scale out with other kinds of infrastructures including like Starlink and so on and SpaceX and stuff.

[00:14:39]So that is going to actually determine which of these companies, you know, you see OpenAI, Anthropic, Google, right?

[00:14:45]The they're all working on their language models. Again, all built on our data, which to me is the big big issue um because of the economic insecurities tied to these issues. But at the same time, it really whichever bet ends up panning out and materializing across industries is likely to produce the most, you know, positive outcome for investors. So why not diversify from the investor perspective at least across these AI companies?

[00:15:08]>> Yeah. And the other thing Dan is that when we look at corporate governance especially you know in the tech times or tech bubble times if you like they're always two steps behind the companies.

[00:15:19]Now how does corporate governance actually get ahead of the curve or is it something that is even warranted look you're betting on a lot of times you know the CEO. I mean if you look for Musk you're betting on Musk with SpaceX.

[00:15:35]You're betting on the godfather of AI, Jensen, when it comes to Nvidia. You're betting on Zuckerberg with Meta. I mean, you know, I think that's how this is going to continue to play out. Look, and are there, you know, in terms of voting rights, corporate governance issues, you know, obviously some could differ, but the reality is investors are focused on the technology. They're focused on what the growth looks like. And a lot of times you're betting on these founders and on these CEOs, but investors, they're not forced to. Now you could say with indexes and what's going to happen they'll obviously have exposure but it's in my view this is nothing like ' 08 nothing like 99 in terms of what I saw you will have froth but overall I continue to think there's a tact bull market next few years >> Adrian so the old adage it's different this time right >> yeah I mean I think it is different in back in 1999 you had sort of hundreds of companies coming through this process.

[00:16:36]You know, names like pets.com that famously kind of popped up and then and then went bust. Um here you've got a very few mega corporations. And you know, when Dan says you can take bets on these founders, you can. I mean, Musk has been incredibly successful in some of his businesses. He's been incredibly unsuccessful in things like XAI and what he did to that. Look at Mark Zuckerberg, incredibly successful in acquiring WhatsApp, Instagram, and building out Facebook. Terrible when it comes to something like the metaverse where he burnt huge amounts of Facebook cash on something that ended up being a complete dead end. So, you're betting on people who in a sense don't have to live with the consequences of failure. And I think that for investors is what makes people nervous because if you look at the current structure of some of these share deals, you know, there's no ability for any kind of activist uh participation.

[00:17:31]You know, the old idea of public markets was you didn't like what the CEO was doing or what the direction of the company was. You could get together with other investors and you could force some kind of change of direction. That's not going to happen with these kind of share issues. And I think that's a problem for people who are having their pensions or their funds through ETFs and things like that thrown into these kind of things is no one's looking out for them ultimately. And if the CEOs, charismatic and brilliant as some of them are, make really silly bets, then there's no payback for the ultimate person who's holding the financial risk, which is the investor.

[00:18:08]>> What are your thoughts, Romesh, on the whole regulatory issue?

[00:18:12]Well, where is that? We've seen the past I said that only a little bit tongue and cheek. We've seen the passage of some regulatory action for example in the European Union. We've seen next to nothing here in the United States. Um you know we've actually um tried to work with different state assemblies to try to have some basic consumerf facing regulation and also to try to think about where work is going to go including some of the data issues right data of ours is being taken you know used to train these models and so on. um natural resources, water, energy, etc. are being in put into all of this, but this is creating potential higher utility costs. So, there are all these different kinds of concerns we're trying to implement. But, uh Trump did uh push forward an executive order that made it a little more difficult for states to engage in regulatory action that's sort of simple and basic common sense, you know. Um, I also believe that uh AI generated content, at least viral AI generated content should at least be somehow watermarked so we know to some extent what we're looking at online. Um, but as as both guests have mentioned um that we're at we're at this time where what we call corporate governance basically means um trusting in um at times the wayward actions and directions of some of these technology CEOs. And so that's just an incredibly small number of people for a small number of companies that are making huge decisions about a technology that's in many ways defining and shaping um our lives.

[00:19:39]>> Romesh again you know then let's look at the actual practicalities and uh actually what's going on in terms of the the ground uh here as well. Are we seeing a productivity revolution or a productivity illusion?

[00:19:54]>> Well really depends where where you look. I I do see you know uh some some use cases where these AI technologies are augmenting people's work right that's really was the idea that was the idea uh around AI in many cases you know even if we date back decades to the history of AI was it would be technologies that would augment or expand people's activity their productivity their creativity could could complement these technologies right um but at the same time I just don't see that many examples certainly myself as a professor in a university um you know I have my students sign an AI pledge. That's the best we could do to kind of ensure that they're actually reading and writing because in many ways what you we're seeing are technologies that are reading, writing and in a sense thinking and essentially speaking for us and uh that's a big big issue that we're seeing at play. So we have to organize on the ground to um re-imagine what work and worker life uh will look like, what economic life will look like and what being um a student, teacher or just you know a citizen ideally in a democracy will look like moving forward.

[00:20:59]>> Um Dan, I'm going to reframe that same question to you because we've got you know fans of AI saying that products from AI are exceeding supply and uh driving major improvements in efficiency. But on the other side of the coin, you got companies which are really at the moment reporting very little productivity gains here as well. How do you reconcile these two realities?

[00:21:22]>> I think the 700 billion that's being spent this year and by a trillion next year, look, it's about driving the use cases and productivity. I think you've started to see that Palunteer, Snowflake, Data Dog from a software perspective. But the use cases are going to play on the enterprise and ultimately on the consumer side with Apple and others. But that's the prove me period.

[00:21:43]In other words, the next 6 to 12 months to to both guest and they make great points. Now, you need to execute show monetization. It can't just be a capex boom for the coming years. you need to see monetization and that will determine ultimately who the winners are, who the losers are. But when it comes to like jobs and what I almost view as a tech PR problem right now relative to to some of the scare tactics around and you're seeing it and I'm sure Romesh sees it every day from a college perspective. I do think many jobs would be created over the years, but right now we're definitely going through one of these transition periods. That's the best way to characterize it. Well, a transition to what? And Adrian, so what's your biggest fear for AI and what's your biggest hope for the technology?

[00:22:31]>> Well, look, I mean, the biggest hope is that it is going to transform the way we organize ourselves. Um, you know, just to go back in time, right? Pets.com was going to sell you uh pet food over the internet. Spectacular bust uh in the first kind of big.com bubble. Well, the guy who's about to try and take over eBay from GameStop made all his money selling online pet food. So, the model does actually work. You know, that was 15, 20 years after Pets.com, right? But it's not Pets.com who are making that money. It's somebody else with a new idea and a new model. And I think what we're going to see is it's going to be new companies who solve this. Old companies are typically terrible at implementing new technology. You can see that with some competition recently to spend tokens, you know, being rewarded at Meta and then they discovered this was a great way to burn cash. You know, companies typically big companies are very bad at implementing smart new technologies. We saw that in the first kind of version of the internet uh when you know people like the Washington Post for example, brilliant in digital publishing had no idea how to make money out of it. You know that's the kind of you know the hope is that new emerging organizations and new emerging companies are going to come in take this technology and really do something with it. The problem is right now we don't know who they are and investors don't really know who they are and we probably haven't seen very many of them. And that's the kind of fear and concern. And also the very real fear and concern that it could be your job that's being transformed without any knowledge of what it is you're going to be doing in 5 years time to earn a living and pay your mortgage and keep your kids in school.

[00:24:17]>> Well, I'm not sure I'll be doing this job if AI continues the way it is. Now, tell me something, Dan. Um, is this the greatest economic opportunity uh of our generation or the most expensive leap of faith in history?

[00:24:31]I think it's I think it's it's a fourth industrial revolution. I think it's this is truly it's a once in a hundred year or plus you know type of opportunity.

[00:24:43]And again I don't just mean from an investment perspective. I mean what we're seeing in terms of the technology in this is the future of robotics autonomous space going to Mars. I mean these are some of the things that we're talking about and I think it's a super exciting time. I get some of the fears. And the problem too is that a lot of bears, they live in hibernation mode in the caves.

[00:25:07]Have they and they've missed every transformationational text the last 25 years.

[00:25:12]>> Romesh, same question to you. Um, and just tie it in with hopes and fears.

[00:25:21]>> Great hopes for uh this technology to expand and augment um what we do, you know, as human beings first and foremost.

[00:25:28]um the kind of work we can do. Uh great hopes for this technology to be actually applied to areas like drug discovery uh to areas like climate change and certainly to the possibility that smaller businesses other kinds of institutions and entities um can emerge which actually create the type of employment um that we need. So absolutely alarmism certainly not doom's you know terminator scenario around AI these are all nonsensical the key is to understand this technology for what it is and I would I would bet that the vast majority of us um around the world do not really understand how this technology how these technologies work how they function I'm concerned that that's somewhat a design choice by these companies of course they have proprietary models I understand that but I think for us to move forward with this we need to create the kinds of institutions can that can quickly adapt to these technologies to ensure that people are protected around the world.

[00:26:24]And that's really my concern is that that's not happening. That's that's going to take global agreements and global innovation to emerge. And I'm just concerned that uh some of the concerns we're seeing with the tech economy where smaller and medium-sized businesses that create employment are not necessarily flourishing partly because of these, you know, large behemoth companies which aren't great at creating jobs and we're seeing them cutting jobs in many cases. Um we need to see that kind of those those sorts of entities emerge and they have to be cultivated right away.

[00:26:56]>> Romesh Shini Vasan, professor at the UCLA department of information studies and specialist on AI and tech. Daniel Ives, global head of technology research at Wedbush Securities and Adrien Monk, senior adviser on AI and technology to the United Nations and editor of the seven things newsletter. Thank you all for joining us.

[00:27:17]And that's it for our show. Get in touch with us on xrishard TV and do use the hashtag ajctc when you do or drop us an email counting the cost at aljazer.net is our address. But there's more for you online at aljazer.com/ctc.

[00:27:33]That'll take you straight to our page which has individual reports, links and entire episodes for you to catch up on.

[00:27:40]Well, that's it for this edition of Counting the Cost. I'm Rashad Salamoth and from the whole team here in Doha, thank you for joining us. The news on Al Jazzer is next.