The Anti-AI Movement and the Myth of Craftsmanship
Why Kitchen Purgatory is the real threat to software engineering, not the machines.
As the CTO of a company whose entire mission is to automate complex business operations with AI, I spend a lot of time thinking about how this all affects labor. Given I have been building software for nearly 30 years, I find the whole “anti-AI” movement especially curious. But before getting to the serious part, I need to talk about salad.
When my office was in our first location, I used to regularly eat at a nearby salad chain that ran on people plus the usual kitchen gear—mixers, ovens, dishwashers. Nothing exotic. What was exotic was the experience. Mondays, Wednesdays, Fridays: your order was only ten or fifteen minutes late, which practically counted as “on time.” Some Fridays they’d mark your order as ready… which meant “ready for you to stand here for an hour while we pretend everything is fine.”
Ingredient accuracy was equally adventurous. Every week something important went missing—occasionally the greens, which is frankly an impressive oversight for a salad company. Sometimes bonus ingredients would appear, like a surprise side quest. A few even triggered my allergies, turning lunch into a negotiation and an extra wait. To be clear, the salads tasted good; the problem was the chaos surrounding them.
Then we moved to a new location.
The shop there is still large, but the production line looks like something from a mild-mannered sci-fi movie: industrial machines that assemble salads quickly, consistently, and without drama. The human team is smaller, but noticeably friendlier and less stressed, probably because they’re no longer running a daily reenactment of “Kitchen Purgatory.” Their job is to keep the machines stocked and clean; the machines handle the precision and speed. Since moving, I haven’t had a single missing ingredient or allergy-inducing surprise.
The contrast between those two locations gets at something important. Automation didn’t erase the humans—it changed what the humans were doing. It took away the part of the job that was frantic, error-prone, and thankless, and left the part that requires judgment, care, and actual interaction. The result wasn’t dystopia; it was calmer staff, happier customers, and a product that finally behaved like the product.
This is the part that gets lost in the anti-AI discourse in software development. People imagine automation as a giant metal arm yanking keyboards away from engineers. In reality, most of what AI touches first are the digital equivalents of repetitive chopping and portioning: writing boilerplate, chasing down syntax errors, or manual regression testing. The work that burns people out. The tasks everyone quietly hopes someone else will pick up.
The writer Ted Chiang once famously said that “fears about technology are fears about capitalism”—the worry that efficiency gains will only ever be used to squeeze more out of a smaller workforce. It’s a valid concern at the macro level, but it ignores the micro level: the quality of the work day itself.
The anti-AI movement treats all automation as a threat to craftsmanship, when the real threat to craftsmanship is doing everything manually while drowning in a backlog. It’s the “Kitchen Purgatory” version of software: stressed teams, unpredictable delivery, and constant rework because something fundamental got missed. In the automated model, the goal isn’t just to do more with less; it’s to create an environment with fewer bottlenecks and more focus on the parts that actually require human skill.
When we remove the “digital chopping,” we don’t just get more code; we get calmer engineers who can actually think about the architecture, the security, and the user.
Change always feels like loss before it feels like relief. But the salad machines didn’t take jobs away—they took pressure away. And in software, as in lunch, that’s usually where the improvement starts.