AI is changing how I think about work

I have been using AI tools seriously for about two years now.

Not as a replacement for thinking, but as something closer to a fast junior collaborator who is very confident, sometimes wrong, and occasionally brilliant.

What actually changed

The biggest change is speed.

Ideas I would have sat on for weeks because implementation felt too slow are now testable within days. Drafts that would have taken hours are now starting points.

This sounds like pure positive. It mostly is. But it also creates a new kind of discipline problem: moving fast without thinking carefully enough.

The risk I watch for

AI tools are very good at producing plausible outputs. The problem with plausible is that it is not the same as correct, honest, or right for your specific situation.

I have caught myself using AI to generate text that sounds like me but is subtly not quite me. The voice is close but not accurate. The ideas are reasonable but not exactly what I would have said.

This matters more in some contexts than others. For code and research, plausible is often fine. For writing about your own thinking, it requires much more careful curation.

What I use AI for

Research drafts. Code starting points. Quick summaries. Brainstorming. Editing suggestions. Formatting tasks.

What I do not outsource: the actual thinking, the real opinions, the honest reflection. That has to stay mine or the whole thing becomes hollow.

Why this is interesting

We are in the early phase of figuring out what AI is actually good for versus what it seems good for.

The people who will get the most from this technology are the ones who are careful about which parts of their thinking they protect and which parts they are happy to accelerate.

I am still figuring out that line.