What Happens to a Hobby When the Hard Part Becomes Optional

Ines started watercolor at sixty-one, the week after she retired. For the first four months she produced what she cheerfully calls "puddles with ambitions." Then, slowly, the puddles became harbors and hillsides. She can point to the exact painting where she finally understood how to leave white paper for light.

Last spring her nephew showed her an app that turns a photo into a convincing watercolor in about eight seconds. She tried it on a picture of her garden. The result was better than anything she'd painted.

"I didn't paint for two weeks," she told me. "Then I realized I'd been asking the wrong question. I kept asking whether my paintings were good. The question was whether painting was good. For me."

Ines is a composite of several readers who wrote to us this year, but her two weeks of not painting are very real and very common. Generative tools have made the hard part of many hobbies optional. That's forcing a lot of people to figure out what their hobby was for.

The hard part was the point

Most hobbies have a difficult core: the part that takes months to learn and gives you most of your frustration. Getting the wash right in watercolor. Fretting chords cleanly on guitar. Making a crossword grid where every crossing works. Writing a short story that doesn't fall apart in the middle.

For a long time, that difficulty was a gate. You had to get through it to get the result. Now, for many crafts, there's a side door. You can get a passable result without walking through the gate at all.

That's a gift if you only wanted the result. But many people discover, the moment the side door opens, that the gate was the part they loved. The struggle was the hobby. The painting was a souvenir.

Three kinds of hobbyist

After talking with dozens of painters, home cooks, amateur coders, musicians and knitters, I've come to think people fall roughly into three groups, and AI affects each one differently.

Makers love the process. The feel of the brush, the sound of the chord coming together, the problem-solving of a stubborn bug. For makers, AI is mostly irrelevant to the hobby itself — you can't automate the pleasure of doing something with your own hands. Some makers use AI as a teacher: "Why does my wash keep blooming?" That's a good use.

Finishers love the result. They want the birthday cake that looks like a castle, the custom song for the anniversary, the working app. For finishers, AI is a huge accelerant, and there's nothing wrong with that. A finisher who uses an image tool to design a tattoo concept isn't cheating at a hobby; they never had that hobby.

Explorers love novelty. They try a new craft every season. For explorers, AI is a mixed blessing: it lowers the barrier to trying things, but it can also let them skip the stage where a hobby becomes theirs.

Most of us are a mix, and the mix varies by hobby. I'm a maker with watercolor, a finisher with home decorating and an explorer with everything else.

Scaffold, not substitute

The most useful distinction I heard came from a hobbyist woodworker I'll call Tobiah, who builds chairs in his garage. "There's a difference between a jig and a robot," he said. A jig is a guide that holds the wood steady so your cut is straight. You still make the cut. A robot makes the cut for you.

AI can be either. Used as a jig, it's a patient tutor, a reference librarian and a sparring partner:

• A guitarist asks for three exercises targeting the specific chord change that keeps tripping her up.
• A home cook pastes in a failed bread recipe and asks for likely causes.
• A novelist asks for blunt feedback on a chapter, then rewrites it herself.

Used as a robot, it produces the thing itself: the song, the image, the story. Sometimes that's exactly what you want. But if you're a maker, it's worth noticing which one you're reaching for.

What we lose if everyone takes the side door

There's a broader worry here, and it isn't about any one person's hobby. Crafts survive because people go through the gate. Today's beginner watercolorist is tomorrow's teacher, and the day after's innovator who finds a new technique nobody had tried. If fewer people learn the hard part, fewer people will be able to teach it, and the craft gets thinner.

I don't think that's inevitable. Photography didn't kill painting; it freed painting from having to be a record of how things looked, and painting got stranger and more interesting. Something similar may happen now. When a machine can make a technically competent watercolor, a human watercolor becomes valuable for different reasons — its quirks, its intent, the evidence of a specific hand.

Ines, again

Ines is painting again, three mornings a week. She uses the app occasionally, to see how a scene might work before she starts. She calls it "a sketch I didn't have to make."

But she keeps her favorite painting — an early harbor, lopsided, with far too much blue — pinned above her desk. "It's terrible," she says, delighted. "And it's mine. The app has never made anything that's mine."

Ines, Tobiah and the other hobbyists in this piece are composites, with invented names and details.