The AI efficiency paradox

I normally like to write about coding challenges and the solutions I found along the way. This time, I just want to get something off my chest. To whomever it may concern.

AI lets me get so much more done as a developer. That’s exciting, and it’s also starting to feel a little exhausting.

I’d expect getting more done in less time to leave me with more free time. Finish what I wanted to do, close the laptop, enjoy the rest of the evening. Seems reasonable.

Somehow, I end up with less.

There’s always time for one more thing

The problem is that being more efficient makes spending more time feel easier to justify. If I can get this much done in an hour, imagine what I could do with another hour. It feels like such a good use of my time that stopping almost seems like a waste.

An idea that might once have been too much work to start on a weekday evening suddenly feels within reach. So I start. And once it’s taking shape, there’s another thing I want to try. Something to improve. Something else that suddenly seems possible.

I often find myself vibe coding till late in the night.

Evenings, weekends—the time I might have expected to get back is going straight into more coding. Each extra bit feels small enough to justify on its own. Added together, they take up quite a lot of time.

And I enjoy it. That’s part of what makes this difficult. Building something and seeing it work is fun. Being able to do that faster makes it very easy to keep going. I don’t need someone telling me to work longer when I can happily talk myself into it.

The amount I want to do keeps growing along with the amount I can do. Finishing something gives me room to start the next thing, and I keep filling that room.

Getting more done, falling further behind

Then there’s the feeling of trying to keep up.

If AI helps other developers produce more too, there’s potentially more of everything to look at: projects, tools, releases, ideas. I can move faster on my own work and still feel like I’m falling behind everything happening around me. I suspect people in other disciplines might recognise some version of this too.

Being able to generate more doesn’t automatically give me more time to read it, understand it, or decide whether it’s any good. Those things still ask for my attention. And there’s only so much of that to go around.

So there are two things pulling at me. More things I feel I should keep up with, and more things I’m excited to build myself. AI helps me do more, but it also gives me more reasons to stay at the keyboard.

I don’t have a neat solution to end this article with. I’m still trying to figure out how to enjoy all of this without giving it every spare hour. Apparently, getting better at using my time and getting better at deciding when to stop are separate skills.

Despite all that, I still believe we live in an exciting era. I think we’ll see great things happening in our lifetime.

Or perhaps we’ll witness the end of humanity. Either way, I’m here for it.