Yann LeCun, former chief AI scientist at Meta and founder of AMI Labs, said in a recent interview that Elon Musk’s artificial intelligence startup xAI is a “failure” and cannot compete on the frontier of artificial intelligence. He also laid out his thoughts on what could lead to a "bubble bursting" in the industry.

Yang's comments reignited a long-standing dispute between him and Musk and cast doubt on the valuations of some of the world's largest artificial intelligence companies.

Over the past few years, Yang and Musk have disagreed on a number of topics, including artificial intelligence and what he calls "conspiracy theories" spread by Tesla's CEO on social media. Musk accused Yang Likun of "knowing nothing about artificial intelligence for a long time."

Yang Likun said in a recent interview: "Frankly speaking, xAI has failed to a certain extent because the founding team has been disbanded.

"Elon is in a very difficult position right now, and it's hard to recruit top talent in artificial intelligence because he didn't treat his previous team well."

Several of xAI’s co-founders have left the company over the past year. In February this year, Musk merged SpaceX with xAI. This huge deal valued xAI at US$1.25 trillion.

In the first quarter of this year, SpaceX's artificial intelligence unit (including xAI) recorded an operating loss of $2.5 billion. Meanwhile, Likun Yang’s AMI Labs raised $1 billion in a March funding round to build a global model, giving it a pre-money valuation of $3.5 billion.

Yang Likun said that xAI has a "huge infrastructure" and leases it to other companies "because that is the only way for him (Musk) to recoup costs."

The infrastructure Yang Likun refers to is xAI's Colossus 1 and Colossus 2 data centers in Memphis, Tennessee. Both Google and Anthropic rent computing power from xAI data centers.

"I'm not very optimistic about the prospects of xAI," Likun Yang said, adding that he didn't think xAI could compete with the two giants, OpenAI and Anthropic.

'The huge bubble bursts'

Corporate spending on artificial intelligence has come under scrutiny in recent months as the technology has cost far more than expected. OpenAI CEO Sam Altman reportedly said during a company livestream this month that companies are currently discussing their spending on artificial intelligence. Altman called the cost of AI a "huge problem."

Yang Likun said: "The price of artificial intelligence services is rising, and while the operating costs are falling, they are not falling fast enough. So, all these companies are losing money, and basically, most people's use is funded by investors. This can't last long, right?"

He added that labs like OpenAI and Anthropic "either have to raise prices or cut costs, or there's going to be a huge bubble burst."

Yang Likun has long been critical of the limitations of large language models (LLMs), which are the basis of current machine learning. Yang Likun even advocates the "world model".

Solving the problem of overinvestment in artificial intelligence is the challenge facing OpenAI and Anthropico

LLMs learn language patterns to predict what will come next, which makes them ideal for inference and programming. Yang Likun's World Model takes a different approach, focusing on building an understanding of how the real or simulated world works. This involves objects, causation, and behavior.

“I personally believe that only agent systems based on world models can be truly universal and reliable,” Yang Likun said.

From Anthropic to OpenAI, artificial intelligence companies are focusing on artificial intelligence agents (AI agents), systems that can perform more complex tasks autonomously.

Yang Likun said that LLM is very useful in fields such as programming or mathematics. But he noted that "the cost of running such a high-performance system is very high compared to what users are willing to pay."