Zuckerberg admitted to his colleagues: ‘The development of AI agents is not going as well as we had hoped…’

It seems that replacing people with artificial intelligence is not such a simple task, at least judging by the example of Meta. In an internal meeting with employees, CEO Mark Zuckerberg admitted that the development of so-called AI agents is not progressing as quickly as he had hoped, despiteč huge investments and a major restructuring of the company, which included thousands of layoffs.

A rare recognition at an internal meeting

At a meeting held in early July, Zuckerberg addressed staff and gave a surprisingly candid assessment of the current state of Meta’s AI ambitions. He admitted that they had hoped for faster progress, especially after the company diverted significant resources and manpower in the direction of artificial intelligence. His statement resonated among employees who witnessed turbulent changes within the company in recent months.

​- The trajectory of the development of agents in the last čfour months has not accelerated in the way we expected – said Zuckerberg, as reported by Reuters.

In addition to the development delays, Zuckerberg also addressed the recent layoffs, admitting that the process was not “as smooth” as it should have been. He added that top people at the company “were concerned that we were not moving fast enough to adapt” to the changing environment in the technology industry, which was the main driver. painful cuts.

Billions invested in AI and painful layoffs

This recognition comes after Meta implemented drastic measures to fully focus on artificial intelligence. Earlier this year, the company laid off about eight thousand employees, which is roughly ten percent of its corporate workforce. At the same time, another seven thousand people were moved to various AI teams, including a department called “Agent Transformation”, Bloomberg reported. All this is part of a wider restructuring aimed at financing expensive investments in AI infrastructure, which is expected to reach as much as 145 billion dollars this year.

Zuckerberg acknowledged that the perceived benefit of the company’s new AI-focused structure “hasn’t materialized yet.” Nevertheless, he expressed his belief that the company will begin to see more significant benefits of its investments in the next three to six months.

Difficult conditions for engineers

While management expresses optimism for the future, morale among some employees appears to be low. Several investigative reports have described Meta’s new AI department as a “soul-destroying gulag,” according to the testimony of some engineers who were transferred there. These allegations cast a shadow over the company’s big ambitions and reveal the human cost of the technological transition.

At the same meeting, CTO Andrew Bosworth tried to calm employee concerns about controversial mouse-tracking software. He explained that an internal investigation found no evidence that employee data was used to train AI models and that the program, which is currently on pause, would only be available on a voluntary basis if it were to restart.

The challenges are not unique to Meta

Zuckerberg’s comments provide a rare glimpse into the broader challenges facing the entire tech industry in the race to develop autonomous AI agents. Although companies are investing hundreds of billions of dollars and launching new products, such as Meta’s “Meta Business Agent” tool for businesses on Instagram and WhatsApp, actual progress toward truly autonomous systems has been slower than public expectations.

The problem of insufficiently used resources is also not alien to others; for example, SpaceX also faced an excess of computing power built for AI models and is considering renting its capacities. These examples highlight the widening gap between the huge capital investment in artificial intelligence and the actual reliability and capability of current systems. Analysts, such as Holger Mueller of Constellation Research, believe that Zuckerberg’s candid comments are ultimately good for the company because they provide employees with a clear understanding of the current state.

*with the use of AI