By 11pm, my brain is done. My AI agents can keep working.

For more than 30 years, going to bed meant the programming stopped. An idea would wait until I could give it a few hours. Those hours had to come from client work, family, or sleep.

So I filtered ideas before I tried them. Would this be useful? Would it work? Was finding out worth giving up an evening?

A lot of ideas never survived those questions.

Now I can hand a feature to an agent before bed and have work waiting for me in the morning. I still need to review it. Sometimes it needs another attempt. It isn't free, and I still have to pay attention. But I don't have to spend the night writing the software.

That changes how much uncertainty I'm willing to accept.

I can ask an agent to prototype a tool I'm only half sure I want. Have it try three approaches to a feature. Explore an idea that would otherwise sit in my notes until I forgot why I liked it.

If I don't like the result, I can throw it away. The experiment can fail without taking my evening with it.

The last nine months have been dramatic and exhilarating. I'm building far more than I could by writing every line myself.

Now imagine that change coming to the physical world.

I grew up watching The Jetsons, so I've been waiting for a household robot for a long time. I don't know when we'll get one that can reliably handle the variety of work in an ordinary home, or what it will cost. A broken lamp is harder to undo than a bad code change.

But suppose we get there.

The first requests are easy to imagine. Clean up after a messy weekend. Prep tomorrow's dinner. Rake the leaves I can never get my son to rake.

Then the requests get more interesting.

Move the couch to the other wall while I'm at work. If I don't like it when I get home, move it back. Refinish one of the old chairs in the garage so I can decide whether to do the other five.

Lay out the treehouse with stakes and string so I can stand in the yard with my son and see whether we like it there. Try another spot if we don't.

Today, even finding out what I want can consume a Saturday. I have to want the result badly enough to spend my time and my back on the experiment.

If trying became cheap enough, I'd try more. I'd get some things wrong. I'd also finish projects I'd never have started.

That's what has already happened to me with software. My list of ideas has grown because more of them have become reasonable to attempt.

I expect the same would happen at home. Eventually, the treehouse might graduate from something we talk about to something we actually build.

Building starts after breakfast. Then I'm going back to bed.

Two things I'd genuinely like back from you. First, hit Reply and tell me which part of this matches what you're seeing. Second, if there's someone in your world who should be on this list, forward it to them.

- Robbie