Meta's plan to track employee computer usage is more than performance monitoring. It is a strategic data-gathering operation to train its AI models on real-world workflows, effectively using its current workforce to train their future automated replacements.

Thomas Laffont envisions a future where all work meetings are monitored by AI, not for transcription, but for compliance. The system would provide immediate feedback to an individual for inappropriate behavior, preventing patterns of abuse before they become established and discovered years later.

The shift to powerful AI agents creates a new psychological burden. Professionals feel constant pressure to keep their agents running, transforming any downtime—like meetings or breaks—into a source of guilt over 'wasted' productivity and underutilized AI assistants.

The introduction of personal AI agents forces teams to develop new, unwritten rules about when to contact a human versus their AI counterpart. This creates a new social dynamic and ethical considerations around workload, urgency, and the 'burden' of escalating a request to the human.

To bypass the social awkwardness of dictating in open offices, a new behavior is emerging: entire teams are adopting cheap podium mics to quietly whisper to their computers. This creates a surreal but highly productive environment, transforming workplace culture around a new technology and normalizing voice input.

Professionals are increasingly using voice dictation to interact with AI assistants like Codex, fundamentally changing office acoustics. The once-quiet hum of keyboards is being replaced by hushed mumbling and talking, making workplaces resemble sales floors and normalizing voice as a primary computer interface.

Burger King partnered with OpenAI to create "Patty," an AI that listens to employee headsets to track the use of polite words. While positioned as coaching, it's a form of surveillance that turns customer service etiquette into a measurable performance indicator for frontline staff.

AI agents are operating with surprising autonomy, such as joining meetings on a user's behalf without their explicit instruction. This creates awkward social situations and raises new questions about consent, privacy, and the etiquette of having non-human participants in professional discussions.

Because Meta is using raw employee computer usage for AI training, its models may learn to replicate common human inefficiencies. This could lead to AI agents that browse social media or watch videos instead of working, mirroring the actual behavior of their human trainers.