AI Didn’t Make You More Productive. It Made You Worse at Saying No.
Bodil Biering
Someone in the industry made a point recently that’s worth sitting with: most people are not dramatically more productive with AI. There are outliers, sure - people claiming 10x. But en masse, looking at people around them, the productivity gain isn’t what the hype suggests.
What AI has actually done is quieter and more dangerous than that.
It removed one of the most valuable disciplines a team can have: deciding not to do something.
Before AI, if you thought of a feature, there was a natural filter. It was too big, too costly, too much effort for the payoff - so you didn’t build it. That friction wasn’t a bug in the process. It was doing useful work. It was the thing that stopped bad ideas from becoming real ones.
Now the friction is gone. The feature that was “too big to bother with“ suddenly feels easy. So you start it. You get into the feedback loop - tweak, generate, refine, tweak again - and somewhere in that loop you’re moving fast, shipping code, feeling productive.
Except the whole time, the honest answer was: this is a stupid idea anyway. You should never have even started.
That’s the trap. Not that AI makes you build the wrong thing slowly. It’s that AI makes the wrong thing feel cheap enough to attempt at all - and once you’re in motion, “should we be doing this“ rarely gets asked again.
The fix isn’t a new process or a new tool. It’s remembering that saying no to a feature was never about lacking the ability to build it. It was a judgement call. AI hasn’t improved that judgement. It’s just made it easier to skip.
