AI gave you your time back. Where did it go?

Everyone was promised the same thing: AI would hand us our hours back. Instead, most people are working more. The studies are blunt about it. Two out of three people who picked up AI last year ended the year working longer, not shorter. The freed-up time never came home. It got spent.

I feel it too. I run businesses and I am hands-on across a lot of projects, and the pressure is real. But here is the part the headlines get wrong. They tell you leadership is quietly pocketing your time savings and pouring them into more work. Some of that is true. Most of what I see is quieter and closer to home: the overwork is coming from us.

It is partly self-inflicted, and partly the race

Give a person real agency and then hand them AI, and they do not sit back. They push. I have the power to execute so much now, so I execute more. That is human, and it is a good instinct pointed in a punishing direction.

The other engine is competition. Everyone around you is suddenly moving faster, so the whole field speeds up, and that lights a fire under every company to match the pace. Nobody sent a memo demanding more. The bar moved on its own, and everyone started sprinting to stay on it.

So before you blame your boss, look at what the tool did to your own appetite.

The founder trap: over-engineering

Here is the mistake I most want the founders I work with to avoid. AI gives you the capability to build almost anything, so founders who have never shipped product before will now engineer a product to death before a single customer ever sees it.

Think about where AI actually saves you: on analysis, and on execution, the code. That is real time back. The question is what you do with it. Too many founders take that reclaimed time and pour it straight back into more building. More features. More polish.

Distribution is the king of all goals for a founder. It is the thing that decides whether any of the building mattered. If AI just bought you a week, the worst thing you can do is spend it building. Spend it getting the thing in front of people.

The shiny-tool ratchet

It is human nature to chase shiny things. It is exciting, and it makes the work feel worthwhile. AI makes it worse, because it has a bad habit of suggesting a new tool for everything.

I spend a lot of my time reminding teams to keep their eyes on the bigger goal instead. I have watched good engineers abandon a clean custom solution to jump into someone else’s subscription product because it looked easier to onboard, and it backfires more often than not. Every tool you add is another thing to manage. Stack enough of them and you get exactly the exhaustion everyone is describing: busy and fried at the same time, with little to show for it.

”Just get AI to do it”

The most damaging line I keep hearing, from founders about their own teams, is “just get AI to do it,” or “why can’t you finish tonight, AI can handle the workload.”

I get the instinct. But if you have the nerve to say that to an engineer, your own deliverables had better be moving faster too. And here is where it falls apart: you still have to think. You still have to plan and structure the work. Skip that and you will spend eight hours going in circles with AI on something a little strategy would have finished in one. AI does not remove the thinking. It punishes the absence of it.

It also quietly disrespects the craft. There is real talent in how an engineer builds a thing, and “AI can just do it” waves that away, right up until the moment you need it.

The actual secret sauce

The teams getting this right are not using more AI. They are being deliberate about it. The secret is to strategize exactly what you are going to hand AI, and exactly how you want it done, and to have that conversation every single day inside the company.

The opposite, the thing that guarantees the paradox, is delegation by shrug: “you figure it out, use AI, just don’t take up my time.” That is not leadership. New tech is always more painful at the start because you have to front-load the learning, and if the person at the top will not sit in that discomfort with the team, everyone underneath pays for it in hours.

And when you get it right, you do get the time back. The founders who have cracked it use AI to kill meetings, which is close to a universal human wish. They run a note-taker over a long, messy conversation and turn it into a short summary or a small deck that is easier to digest, even for the person who was in the room presenting it. Small, boring, and it hands people their attention back.

Point the tool

The productivity paradox is not really about AI. It is about where you point it, and what you do with what it gives you. Aim it at more building and you will drown. Aim it at distribution, at focus, at clearing the noise so your team can think, and it does what it promised.

The tool was never the problem. The controls are in your hands. Use them on purpose.