The consensus take on AI and time is that it does not actually give you any. Fortune ran the now-standard framing in March: tasks that used to take six hours now take one, workers do not go home early, the work just expands to fill the space.
That read is correct when someone else decides what fills the reclaimed time. When you decide, the same math tells a different story.
I still work the same hours and have won back no leisure. The more interesting change is the tax ambition used to carry in hours. That tax just collapsed.
The tax was mechanical work - the drafting, formatting, summarizing, and chasing, the stitching together of context from six systems before you could write the thing you actually wanted to write. The ratio of mechanical hours to interesting hours was brutal, and it is why most ambitious people run out of energy before they run out of ideas.
Now the mechanical layer runs on rails, and the reclaimed hours shift toward judgment, ambiguity, synthesis, and decisions that still require a person. I am doing more of the work I always wanted to do, at a volume I could not previously sustain.
Here is the part I think is still underestimated: if the tax on ambition was paid in hours, then capability was never the constraint for many people. Capacity was. The scale did not tip toward more ambition or more free time; the floor under ambitious work dropped, so more people get to sustain it, and more of what they already do gets to count.
Fortune is right that AI expands the work. That framing misses how AI also expanded who gets to do it.