Why I Stopped Fearing AI in My Shop

AI is another tool - but the craft is still mine.

 

I often get asked what I think about AI taking over creative work. A few years ago, I would have been one of the people warning other woodworkers about AI. Today, there's a good chance I'm talking to AI more than I talk to anyone else about my shop.

Slowly but surely, AI has become one of the most useful tools in my shop, right up there with my scroll saw and my sander. Here's what I've learned from using it in my shop.

Patterns come together faster - without losing the design work

Before AI, turning an idea into a usable pattern could eat an entire evening: sketching, adjusting proportions, figuring out how a photo reference would actually translate into cuttable shapes. Now I can get from concept to a workable pattern draft in a fraction of that time.

A client once sent me a photo of an intarsia eagle she wanted duplicated. She had no idea who made it or whether a pattern even existed. The original was pretty basic: flat, with little detail or realism. With AI, I built her an entirely original pattern in minutes - not a copy, a genuinely new design, and a far better one. She said yes immediately even though I told her it was created using AI. I had to fix a few things in the design such as where the talons hold the fish but, overall, after building it, that piece is now one of my favourites.

What hasn't changed is the judgment part. AI doesn't know which grain direction will make a wing feel like it's lifting, or where a joint needs simplifying so it can actually be sanded. That's still me. What AI gives me is time. Time I used to spend on the mechanical part of pattern-making, now spent on the parts that actually need my eye.

A second opinion when there's no one else in the shop

Most days, it's just me at the bench. I've got access to a great Facebook community of fellow intarsia and scroll saw artists, and I lean on them, but they're not always around, and sometimes I need feedback in the middle of a project, not three days later in a comment thread.

This is where AI has surprised me the most. I can describe where I'm stuck: a piece that feels flat, a colour transition that isn't working, and get a genuinely useful, honest response back. Not flattery, not a generic "looks great!" Real, constructive input, the kind you'd want from a mentor looking over your shoulder. It doesn't replace the community. It fills the gap between check-ins.

Guitar Legends: A real working relationship

The Guitar Legends series is the clearest example I can point to when it comes to working with AI in the shop. It's a long-running collection, piece after piece, and I've used AI throughout. From research on the instruments and the artists to talking through design choices and troubleshooting the same things I've written about in this blog: finishes, joints, mistakes.

It's not a magic button that makes the pieces for me. It's more like having a knowledgeable collaborator on call, one who remembers what I've already told them about the project and helps me think it through instead of starting from scratch every time. That kind of continuity has made the series better than it would have been had I been working alone.

A real concern: The pattern that was never built

I do want to raise one issue that's not a fear; it's legitimate.

Before AI, if you saw a picture of an intarsia artwork alongside the pattern, you had a reasonable expectation that it had been built and tested. Pattern designers always cut the piece themselves first, to make sure it actually worked, before publishing it alongside the pattern.

That's no longer a safe assumption. AI can generate a beautiful, completely realistic image of an intarsia piece that was never built or tested. I've seen gorgeous work show up on Facebook that, as far as anyone can tell, only exists as a picture. No pattern behind it and no assurance that it can actually be built.

That makes buyers and pattern-users cautious, and honestly, it should. A piece that looks stunning in an image can still be structurally impossible to build cleanly, or have joints that never close right in real wood. The old proof, "here's a photo, so it must work", doesn't hold anymore.

My advice: if you're buying a pattern, ask if it's been built and tested, not just rendered. And if you're posting your own work, make it clear what's a finished piece versus a concept. That distinction matters more now than it used to, and the craft is better off if we're upfront about it.

Originality still matters

There is one uncomfortable side effect to all this that I think is worth acknowledging.

If you create an original intarsia pattern and post a detailed photograph of the finished piece online, someone else may be able to work backwards from that photograph and create a pattern of their own. AI makes that process easier, but it isn't entirely new. A determined person with a good eye, a tracing method, or even a screenshot can already extract a surprising amount of information from a photograph.

That can be discouraging, especially if you're just getting into the craft.

After all, designing an original pattern is one of the most rewarding parts of intarsia. You don't just want to say, "I built this." You want to say, "I imagined this, designed it, figured out how to make it work, and then built it."

I don't think we should lose sight of that.

A photograph can be copied. A pattern can potentially be reconstructed. But that doesn't make the original act of creating it any less valuable.

In fact, I think it makes the distinction between creating and making more important.

AI can help with the first. It can generate ideas, explore possibilities and help turn a concept into something workable. But the experienced artisan still decides what's worth pursuing, what needs to change, what will work with real wood, and what ultimately deserves to exist.

And once the saw starts cutting, there is nowhere for AI to hide!

The grain still has to cooperate. The pieces still have to fit. The joints still have to close. The finish still has to work. And when something goes wrong, somebody still has to figure out why.

That's where the satisfaction of this craft lives.

So if you're new to intarsia, don't let AI convince you that originality no longer matters.

Design something. Make it yours. Be proud of it.

Then build the damn thing!

Why are so many crafters afraid of AI?

I think most of the fear I hear from other makers doesn't come from experience — it comes from a lack of it. People imagine AI as something that can replace them because that's the story that gets told the loudest. Very few of the people with the strongest opinions have actually sat down and used it on a real project.

Fear of the unknown is normal; I get it. I spent almost 40 years in IT, and I've watched plenty of "this will replace us" tools come and go. Most of them didn't replace anyone. They just changed what people spent their time on. I think AI is going to be the same for this craft: a shift in where the hours go, not a replacement for the hands doing the work.

If you're still skeptical, start small

You don't have to overhaul how you work to find out if this is useful to you. Pick one thing, ask AI to help you rough out a pattern idea, or describe a piece you're stuck on and see what it's got to say. No commitment, no big investment.

You don't have to trust AI. You don't even have to like it. But before deciding that it has no place in your workshop, try putting it on the bench beside you. Who knows, you might be pleasantly surprised.