Trillion-Dollar AI Bubble On Verge Of Popping?
Support CleanTechnica's work through a Substack subscription or on Stripe.
The amount of money being invested in AI (artificial intelligence) is wild. So much money is being poured into the industry. Investors all want to bet on leaders of this new era, and large AI companies are pouring a lot of that cash into enormous data centers packed with unbelievable amounts of computer hardware, and powered by dozens upon dozens of dirty, polluting power plants. However, has it all gone too far?
Note that this is not to say AI isn’t doing amazing things and won’t continue to be more and more useful. However, there’s some concern the hype and the investment have gone too far. Let’s roll through a handful of recent comments and facts.
Yann LeCun, one of the “Godfathers of AI,” is one of the notable people who think the industry has been far too overhyped and misunderstood. He’s been pointing out that AI costs could be much higher than the amount of money customers are willing to pay for it.
“The prices are going up of those AI services, but the cost of running them is going down, but not nearly fast enough,” LeCun recently said. “And so all of those companies are losing money, and basically, the use for most people is funded by the investors. That can’t go on for a very long right?”
He argues that the industry will either have to cut costs significantly or raise prices. “Labs like OpenAI and Anthropic are going to have to increase prices, they’re going to have to cut costs, or there’s going to be a big bubble explosion.”
LeCun adds that xAI, now swallowed up by SpaceX, is in particular risk. “xAI is kind of a failure, frankly, because the founding team has [left],” LeCun shares. “Elon is now in a position that is very, very difficult for him to kind of hire top people in AI, because he’s kind of, you know, not behaved in sort of very good ways toward the … previous team,” LeCun added. Then there’s the fact that it scaled up its hardware/data center infrastructure … much more than it actually needed. Its huge amount of computing infrastructure has far exceeded its own actual demand, so it’s now resorted to renting out that digital space to competitors (Anthropic and Google) — “because that’s the only way he [Musk] can recoup the cost.” As we’ve reported previously, xAI had a $2.5 billion net loss in the first quarter of this year alone. With SpaceX now a public company, it’s going to want to turn that around as much as possible ASAP.
But back to other companies, like the ones with more demand that have to rely on xAI’s extra computing infrastructure. They still have the issue of spending far more money than they are bringing in. “Labs like OpenAI and Anthropic are going to have to increase prices, they’re going to have to cut costs, or there’s going to be a big bubble explosion,” LeCun warns. And LeCun doesn’t see the fundamental systems these companies are using (large language models) as good enough, reliable enough, and efficient enough to make that happen. H doesn’t see them as offering enough value to increase pricing significantly, so sees a financially reckoning coming.
Elsewhere, there’s reporting that large corporations spending a ton of money on AI are starting to find that it may not actually be worth the high expenses. “A confusing contradiction is unfolding in companies embracing generative AI tools: while workers are largely following mandates to embrace the technology, few are seeing it create real value,” Harvard Business Review writes. “Consider, for instance, that the number of companies with fully AI-led processes nearly doubled last year, while AI use has likewise doubled at work since 2023. Yet a recent report from the MIT Media Lab found that 95% of organizations see no measurable return on their investment in these technologies. So much activity, so much enthusiasm, so little return. Why?” One answer is that a lot of what’s being created is, as HBR summarizes, “workslop.”
“In their pursuit to boost productivity, become less reliant on human labor, and reassure investors that they’re riding the cutting edge of tech, some nagging issues are cropping up,” Futurism adds, and “over-relying on AI can prove disastrous for organizational knowledge, the critical business insights companies need to make strategic decisions.” Shocker.
Think of it like this: Imagine middle school students are using AI to do their homework and get answers for in-class work. When it comes time to doing such work without the help of AI, would they score as well as if they had done everything in the normal, traditional way? Heck no. Workers at large corporations aren’t going to fare much better when it comes to figuring things out and making decisions after using AI like a crutch.
“The phenomenon, dubbed ‘knowledge decay,’ describes the deterioration of information over time, marked by workers forgetting skills and organizations relying on outdated processes. In the context of AI, it can be a dangerous downward spiral that starts with workers using AI to produce low-quality work, which wastes colleagues’ time, erodes trust, and gradually sloppifies organizational knowledge into worthless soup.” Scale that issue up to whole departments or companies, and you have a serious problem developing.
“It’s an already familiar trend. Even in the early days of the AI boom, experts warned that employees may be spending more time hunting down the many errors being made by unreliable and hallucinating AI tools than if they weren’t using the tech at all. Some companies even resorted to hiring workers specifically to fix AI errors.” Also, as the whole system of reliable information and trust breaks down, the gears of an efficient company become creaky if not rusty and much more time and money is wasted for what used to be normal, seamless work. “Errors compound and pile up,” the Harvard Business Review adds. “Trust in information erodes. People spend more time verifying facts or risk costly and dangerous mistakes. Eventually, people start to lose trust in the processes that they rely on to do their jobs.”
Workslop is becoming a major issue at some companies. Cutting back on AI use may be necessary to get the problem under control. There’s been a strong AI trend in large corporations, but that could swing back as an AI backlash and trend away from it. Just as AI companies are under a lot of pressure to raise prices. Uh oh….
That brings us to a third story. “A wave of selling in tech stocks is starting to reflect doubts over whether the spending boom on artificial intelligence is worth it,” NPR reports today. “The best-known AI-related tech stocks, Nvidia and Google-parent Alphabet, were down for a second day in a row. Among the biggest losers on Tuesday, however, was chipmaker Micron Technology, whose shares plummeted over 13%. These sell-offs sent the tech-heavy Nasdaq index down over 2%.” SpaceX is also down 22% in the past five days.
“The market just continues to oscillate between ‘AI is going to be great and increase productivity and all these companies are going to win’ and ‘AI is a big waste of time and it’s not worth the return on investment at all and this is all one big bubble,'” said Gil Luria, head of technology research at investment firm D.A. Davidson. That sure sounds like the dilemma.
NPR notes that $580 billion has been invested by large corporations into AI in the past year, following $1 trillion invested over the four prior years. That’s according to Stanford University’s AI Index Report. Keep in mind everything earlier in the article about these companies starting to discover that AI is becoming counterproductive and causing efficiency and effectiveness problems, or at least starting to seriously wonder about the net benefit. What if that massive trillion-dollar shift into AI swings the other way?
Getting back to LeCun, he thinks a “big bubble explosion” is coming. He also thinks the way the industry really needs to move forward is with “world models” rather than LLMs. So, his own company, AMI Labs, raised $1 billion in March to work on that. Hmm….
“I personally don’t think we’re going to have generalized reliable agentic systems until they’re based on world models,” LeCun notes. We’ll see.
Sign up for CleanTechnica's Weekly Substack for Zach and Scott's in-depth analyses and high level summaries, sign up for our daily newsletter, and follow us on Google News!
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
Sign up for our daily newsletter for 15 new cleantech stories a day. Or sign up for our weekly one on top stories of the week if daily is too frequent.
CleanTechnica uses affiliate links. See our policy here.
CleanTechnica's Comment Policy
