Yann LeCun characterised Elon Musk’s xAI as a “failure,” asserting that it lacks the capability to compete at the forefront of artificial intelligence. He elaborated on his perspective regarding potential factors that could trigger a “big bubble explosion” within the industry. The remarks made by LeCun revive an ongoing dispute with Musk and raise questions regarding the valuations of several of the largest AI firms globally. LeCun, formerly Meta’s chief AI scientist, has engaged in disputes with Musk over recent years on various subjects, including AI and what he characterised as the Tesla CEO’s “conspiracy theories” on social media. Musk has accused LeCun of being “out of touch with AI for a long time.” LeCun is frequently referred to as a “godfather of AI” due to his pioneering contributions to the discipline. “XAI is kind of a failure, frankly, because the founding team has” departed, LeCun said. “Elon is now in a position that is quite challenging for him to attract top talent in AI, as he has not conducted himself in particularly favourable manners toward the previous team.” Over the past year, a number of xAI’s co-founders have departed from the organization.

In February, Musk merged SpaceX with xAI in a significant deal that valued the company at $1.25 trillion. In the quarter concluding on March 31, SpaceX’s AI division, encompassing xAI, recorded an operational loss of $2.5 billion. Meanwhile, AMI Labs secured $1 billion in a funding round in March aimed at developing world models, resulting in a pre-money valuation of $3.5 billion. LeCun stated that xAI possesses “huge infrastructure” which it leases to other firms, “because that’s the only way he [Musk] can recoup the cost.” LeCun’s remarks regarding infrastructure pertain to xAI’s Colossus 1 and Colossus 2 data centers located in Memphis, Tennessee. Both Google and Anthropic have leased computing power capacity at xAI’s data centers. “I’m not very positive about the prospect of xAI,” LeCun stated, noting that he does not foresee xAI being able to compete with major players OpenAI and Anthropic.

Enterprise spending on AI has faced increased scrutiny in recent months as the technology is proving to be more costly than anticipated. OpenAI CEO Sam Altman reportedly stated this month during a company livestream that businesses are currently evaluating their expenditures on AI. Altman stated that AI costs represent a “huge issue.” LeCun stated, “The prices are going up of those AI services, but the cost of running them is going down, but not nearly fast enough. 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?” The AMI Labs founder remarked that entities such as OpenAI and Anthropic will need to either raise prices, reduce expenses, or face a significant bubble burst. LeCun has expressed strong criticism regarding the constraints of large language models, which serve as the cornerstone for the prevailing generation of prominent AI products.

Instead, LeCun advocates for “world models.” LLMs acquire linguistic patterns to anticipate subsequent elements, rendering them particularly adept at reasoning and coding. World models adopt an alternative methodology by striving to construct a comprehension of the mechanisms underlying the functioning of either the real or simulated world. This encompasses entities, causal relationships, and activities. “I personally don’t think we’re going to have generalised reliable agentic systems until they’re based on world models,” LeCun stated. Artificial intelligence firms, ranging from Anthropic to OpenAI, are concentrating on the development of AI agents—systems designed to autonomously execute more intricate tasks. LeCun stated that LLMs are beneficial in domains including coding and mathematics. He pointed out that the cost of operating these high-performance systems far exceeds what users are willing to pay.