Meta CEO Mark Zuckerberg stated that the company's AI agent development is progressing more slowly than anticipated, and recent structural changes have not yet yielded expected results. He projects benefits within three to six months.
Zuckerberg's comments provide a reality check on the pace of AI agent development, impacting the broader tech industry's significant investments and predictions for blockchain payment adoption by autonomous agents.
Meta CEO Mark Zuckerberg informed employees that the company's artificial intelligence agent development is progressing slower than anticipated, according to reports from an internal town hall. Zuckerberg stated that the company's recent structural changes have not yet yielded the expected results, though he projected some benefits within the next three to six months.
The company is investing significant resources into AI, but the competitive landscape has made progress challenging. Zuckerberg described Meta's journey as one toward "superintelligence," emphasizing the hard work required.
In related news, Meta's Chief Technology Officer Andrew Bosworth announced that the company's AI training program, which previously used employees' keystrokes and mouse movements, will now be opt-in only if it resumes. This program was paused last month following an internal leak and employee backlash over privacy concerns. Bosworth acknowledged that the rollout had negatively impacted morale and trust within the company, though he noted it generated useful data.
This reversal follows other recent adjustments, including offering engineers the option to leave the Applied AI task force. These developments occur in the context of Meta's recent 10% staff layoff, impacting approximately 8,000 employees, and Bosworth's prior warning about historically low employee morale.
Pick the topics you care about. Get only what matters, on your cadence.