Meta CEO Mark Zuckerberg made an unexpected admission that has forced the market to reassess expectations for the company's artificial intelligence development. During a recent conversation with analysts, he confirmed that progress in AI agents has significantly lagged behind initial forecasts.

Reality proved more complex than expectations

Over the past four months, according to Zuckerberg, the pace of development in this area has not only failed to accelerate but has also fallen short of even Meta's conservative plans. This statement came amid widespread enthusiasm around generative AI, where many companies, including Meta's competitors, are actively promoting their agent-based solutions.

Of particular interest is the data on the x402 protocol, which supports the operation of AI agents. According to the Artemis platform, over the past 30 days, the transaction volume through this protocol amounted to only about $2 million. For a company with a market capitalization in the trillions of dollars, this is a drop in the ocean, merely highlighting the current immaturity of the technology.

Analytical perspective

The situation surrounding Meta demonstrates a fundamental problem facing the entire industry: creating truly autonomous AI agents capable of performing complex tasks in real-world conditions has proven to be a significantly more difficult challenge than anticipated. Investors, accustomed to rapid breakthroughs in large language models, must now reassess their time horizons. Personally, I believe that the current slowdown is not a failure but a necessary phase of consolidation, without which it is impossible to build a reliable infrastructure for the mass adoption of agents. Meta, with its vast resources, can afford such a pause, but for smaller players, this signal should serve as a warning: the race for AI agents is just beginning, and the winner will be the one who can endure a marathon, not a sprint.