← All posts

Why "save time" is the wrong way to think about AI

Every new tool gets sold the same way: faster. The printing press was faster than a scribe. The camera was faster than a painter. It's the easiest pitch there is, because it measures the new thing against the old one — and that's exactly what undersells it.

The printing press didn't just copy books faster. It let an idea travel further than any one person could carry it. The camera didn't just paint a portrait faster. It created an entirely new way of seeing a single moment, one painting could never have offered at any speed. The interesting thing was never how quickly the old job got done.

So here's the tell: if a tool only ever helps you do the same things you were already doing, just quicker, you haven't found what it's actually for yet. You're still measuring it by the old medium's yardstick.

The question worth asking isn't "how much time did this save me." It's "what can I attempt now that wasn't worth attempting before." A side project too ambitious for a spare weekend. A second opinion on your own work you never had time to get. An idea you'd have shelved as someday.

We don't build tools to help people work the old way, faster. We build tools that help people find the new way — the thing they wouldn't have tried at all, if it still cost what it used to cost.