Mark Zuckerberg considers "bursting of the AI bubble" possible

Is the AI bubble about to burst? Mark Zuckerberg believes this is absolutely possible. In a new podcast from Vox Media, the head of Meta talks about potential bad investments and opportunities. The latter also includes the fact that even if the bubble bursts, there will be some winners.

The podcast hosts are well-known tech journalists Alex Heath and Ellis Hamburger, who have worked in several start-ups and tech companies. Access will provide insights into the tech world, apparently mainly in Silicon Valley. Mark Zuckerberg is a guest in the first episode. It starts with the newly launched Ray-Ban Meta Display glasses and how Zuckerberg uses them—his otherwise apparently long WhatsApp messages are getting shorter. The fact that he believes glasses will replace smartphones is nothing new.

What is interesting, however, is that Zuckerberg says Meta has only produced a few hundred thousand of the display glasses so far. It is therefore by no means assumed that they will be a mass seller. And at 800 US dollars, it's no wonder. According to Zuckerberg, profit will not be made with the glasses themselves anyway, but with what people develop around them. The aim is to create an ecosystem that benefits Meta.

Meta wants to make a profit from the burst

What Zuckerberg says about AI is also very revealing. That AI is a bubble: possible. But the risk of missing out is greater for Meta. If Meta loses a few hundred billion US dollars, that would be very unfortunate, but the risk on the other side is even greater. “If a company is too slow and an artificial superintelligence emerges earlier than expected, it will fall behind in what I believe is the most important technology that enables most new products, innovation, value creation, and history.”

And finally, a bubble like the dotcom bubble of the 2000s has shown in the past that there are indeed winners—even though many other people could lose a lot of money. However, this is more of an advantage for large companies that can buy up smaller start-ups and what's left over, so to speak.

The promise of AI meets with criticism

The fact that AI has become a bubble has also been criticized by other experts and AI companies in addition to Zuckerberg. Bret Taylor, CEO of OpenAI, and CEO Sam Altman, both recently spoke of a bubble. But they also expect that economic value will nevertheless be generated—as with the Internet. In the case of OpenAI in particular, it is always unclear how the company intends to earn money in the long term. The income is nowhere near enough to cover the expenses. Another 100 billion dollar deal with Nvidia has just been announced. Nvidia is effectively giving OpenAI the money so that OpenAI can invest it in Nvidia hardware.

At the same time, doubts are growing whether Artificial General Intelligence (AGI) could really be achieved as soon as AI companies like to claim. OpenAI, in particular, is constantly setting expectations very high, but GPT-5 is yet another iterative AI model that is slightly better than its predecessor.

Google's DeepMind CEO, Demis Hassabis, has already got carried away and labelled Altman's promises of AI at PhD level as “nonsense.” The term AGI does not even appear at Meta. While a team in the USA is working on an undefined superintelligence, the AI team in Paris has initially set itself the goal of advanced machine intelligence (AMI). Meta's Chief AI Scientist and Turing Prize winner Yann LeCun has repeatedly criticized the Silicon Valley mindset as arrogant and misguided.

(emw)