You’ve heard the promise: AI will make you more productive. You’ll get your evenings back. You’ll finally have time to think. But then Meta’s CTO Andrew Bosworth said something that shattered that fantasy. He told employees that any productivity gains from AI should be used to do more work. Not to work less. Not to go home early. More work.
Let that sink in. The person in charge of one of the world’s most advanced AI operations just told you the truth: the productivity dividend doesn’t go to you. It goes to the company.
The dirty secret of productivity is that it never leads to leisure. It leads to a new baseline.
You’ve probably noticed this pattern. When your team gets a faster tool, your deadlines get tighter. When you automate a task, your manager assigns two more. The goalposts move. The hour saved becomes the hour you’re expected to shovel into the next fire.
But why? Why does every efficiency gain just create more work? The answer isn’t technology. It’s incentives.
One tired commenter on a Business Insider article nailed it: “Does anyone know about the time horizons for executive stock compensation at BigTech? Something has to explain the lack of long-term strategy.”
That commenter understood something most AI hype pieces miss. Executive stock options vest in three to five years. A CEO’s bonus depends on this quarter’s earnings. So when AI makes your team 20% more productive, the rational move for management is not to give you a four-day week. It’s to squeeze that 20% out of you in the form of more output, then up the targets for next quarter. The long-term damage – burnout, turnover, loss of creativity – becomes someone else’s problem.
Your boss’s stock options expire in three years. Your burnout is a Q4 problem.
This is not a technology problem. It’s a power problem. The people who control the productivity dividend don’t benefit from giving it to you. They benefit from extracting it. And AI is the perfect extraction tool – it’s invisible, it’s scalable, and it’s framed as progress.
I’ve seen this firsthand. A friend at a mid-sized tech company watched their team adopt an AI code assistant. Within two months, the team’s output per developer doubled. Did they get to leave at 3 PM? No. They got a new sprint goal: triple the output by next quarter. The AI didn’t create freedom. It created a new minimum.
So what’s the twist? The twist is that the real conflict isn’t AI versus jobs. It’s AI versus you – and the winner is determined by who owns the productivity dividend. If the gains go to shareholders and executives, you’re just a faster horse. If the gains go to workers, you’re actually free.
Next time your CEO talks about AI efficiency, ask yourself: who gets the dividend?
This isn’t a prediction. It’s a description of how the game is already played. Meta’s CTO just confirmed it. The question is whether you’re willing to see it.
FAQ
Q: Isn't AI supposed to make work easier for everyone, not just bosses?
A: In theory, yes. But the distribution of gains depends on power. If management controls the surplus, they'll reinvest it into more work, not leisure. The technology itself is neutral—the incentives are not.
Q: What's the practical takeaway for someone who uses AI at work?
A: Track what happens to your productivity gains. If they result in a higher output target without a reduction in hours, you're being squeezed. The smart move is to negotiate for a share of the dividend—shorter hours, more autonomy, or a bonus tied to efficiency improvements.
Q: Isn't the Meta CTO just being realistic about competition?
A: That's the standard defense. But 'realistic' is a choice. Companies could choose to share gains. The fact that they don't reveals that the real bottleneck is not competition, but the assumption that workers should always do more. That assumption is a political choice, not a law of physics.