Kickstarter Diaries 14: Why Not Use AI?
The latest episode in my tragicomic behind-the-scenes account of life during a crowdfunding campaign.
As the AI debate rages, I started wondering about the Endangered Alphabets initiative that I’ll undertake if my current Kickstarter campaign is successful, and how it would be different with AI.
My plan is to make ten carvings in endangered writing systems and send them back to their communities of origin, where each can be put up in a public place, a place of honour, as an item of pride for the community, an acknowledgement that someone in the outside world knows they’re there, recognizes and respects them, and supports their right to exist, their right to read and write in their own script.
And I imagine someone saying, “Wouldn’t it be a whole lot easier and quicker to make them with AI and a 3D printer? Or use a CNC router, have the laser cut the letters, spray-paint the top black, run it through the sander, and get the same result?”
Let me count the ways that the results might look similar but would be very different.
- The laser cut and even the 3D printer would actually be considerably more exact and consistent than my carved lines, whose slight imperfections would be visible, even if only unconsciously, and would make my version look more human. Someone put effort into this, they will imply.
- The machines can’t feel joy or satisfaction or thoughtfulness or meditation or challenge. If I hand over the act of writing to a machine, then I am robbing myself of those possibilities, and in doing so, I’m sort of robbing the world of a person who has felt those things. I’m a better person for going through all that, and my work is better for being made by a person who has gone through all that.
- If everyone used AI to make endangered alphabets carvings, the number of people who can make a living carving by hand would go down steadily until the profession was abandoned and the hand skills were lost. Moreover, anyone in these communities who has hand skills would start yearning for the days when they can have AI and play with the big dogs.
- Each of these carvings will take me a good six weeks. I’ll spend those six weeks learning by doing—learning about wood, learning about carving, learning about the cultures whose scripts I was carving, learning about writing itself. In several cases I’ll track down people from the script community to ask advice, send out feelers, build bridges. Learning takes time and involvement; ergo all that time and effort are priceless.
- Do I really want to send the message to my friends in minority communities around the world, “I decided that carving some by hand for you was too much work, so I had AI do it?”
Please support old-school endangered alphabets hand carving HERE.
March 22, 2026 @ 11:50 am
Why does the author prefer hand‑carving endangered alphabets instead of using AI and machines like 3D printers or CNC routers, even if they could be more precise?
March 22, 2026 @ 2:49 pm
From my new book By Hand:
In the early days of the Endangered Alphabets Project, I was invited by the Smithsonian to create some kids’ alphabet games for their summer Folklife Festival. One of the games involved rubber stamps of individual letters from endangered alphabets, which were kindly made for me by a neighbor who had a computer-controlled laser cutter.
He liked my handmade carvings, but then observed, “Of course, you do realize that you could email me the design, I could open it in Corel Draw, set it up, export it to my laser, have the laser cut the letters, spray-paint the top black, run it through the sander, and get the same result?”
He’s a good guy, and I’m not sure how serious he was—I think that, as someone high up in IBM and also a woodworker, it just naturally occurred to him to look for a labor-saving approach. And I’m sorry if I’m underlining a point you’ve already understood, but let me count the ways that the two results might look similar but would be very different.
1. The laser cut would actually be considerably more exact and consistent than my carved lines, whose slight imperfections would be visible, even if only unconsciously, and would make my version look more human. Some people would prefer his, some mine, just as some people prefer vinyl recordings, some digital.
2. The computer, the software, the laser, the spray booth and the sander might cost anywhere between $25,000 and $250,000. That sets up quite a financial divide between those who can afford the investment in that equipment and those who have a set of hand tools that cost $70.
3. The machine can’t feel joy or satisfaction or thoughtfulness or meditation or challenge. If I hand over the act of writing to a machine, then I am robbing myself of those possibilities, and in doing so, I’m sort of robbing the world of a person who has felt those things. Again this all comes back to how we define writing. If we define writing as a chore, then it seems reasonable to hand the work over to a robot.
4. If everyone took their endangered alphabets carving jobs to my friend, the number of people who can make a living carving by hand would go down steadily until the profession was abandoned and the hand skills were lost.
5. Each of the particular series of carvings I was doing at the time took me a good six weeks. I spent those six weeks learning by doing—learning about wood, learning about carving, learning about the cultures whose scripts I was carving, learning about writing itself. In several cases I tracked down people from the script community to ask advice, sent out feelers, built bridges. Learning takes time and involvement; ergo all that time and effort were priceless. What my friend learned was almost entirely within the realm of his equipment, apart from one tiny detail: when a laser cuts maple, the basement fills with the smell of toasted wood.