Meta, the parent company of Facebook and Instagram, had been planning to replace a significant portion of its workforce with AI agents. The plan, code-named Project OT, aimed to cut the size of many teams by as much as 60% and have AI take over daily work performed by thousands of employees.
The idea for Project OT was hatched by Mark Zuckerberg and his senior team at his Hawaii compound in January. The plan was to have AI agents perform tasks such as code changes and customer support, with smaller teams of human employees supervising the virtual workers.
However, the plan was ultimately abandoned when the technology did not deliver. Meta's own data showed that the AI agents were producing volume rather than value, with code changes rising 220% year over year, but changes that reached users as new or upgraded features rising only 36%.
The company's employees had also become skeptical of the plan, with many feeling that they were training their replacements. Meta had reassigned engineers to write software puzzles used as training data, and many employees called the work rote in internal posts.
Despite the abandonment of the plan, Meta is still investing heavily in AI. The company plans to invest at least $130 billion in AI chips and infrastructure this year, and has promised no further company-wide layoffs this year.
The incident highlights the challenges of implementing AI in the workplace, and the need for companies to carefully consider the potential consequences of relying on virtual workers. As Meta's experience shows, the technology is not yet ready to replace human employees, and companies must be careful not to overestimate its capabilities.
Meta's decision to abandon the plan is also a reflection of the changing public message around AI. The company is now running advertising saying it is betting on people, and Zuckerberg's recent essay predicts an abundance of jobs, while allowing that individual companies may employ fewer.
The Future of AI at Meta
Despite the setback, Meta is still committed to developing its AI capabilities. The company is working on a range of AI-powered tools and technologies, including a new AI-native playbook that sets out the target structure for its teams.
The playbook calls for product teams of 10 to 20 specialists to shrink to pods of three to five, with engineers, designers, and product managers sharing one title, "builder". Middle management layers will be eliminated, and "agent-assisted analysis" will set daily priorities.
It remains to be seen how successful Meta will be in its AI endeavors, but one thing is clear: the company is committed to making AI a central part of its operations. Whether that will ultimately lead to the replacement of human employees or the creation of new jobs remains to be seen.
Cet article a été rédigé avec l'assistance de l'IA.
News Factory APP - actualités agentiques pour booster votre SEO et AEO.