The author spent 2006 building Flash microsites in Manhattan. Flash was objectively bad — inaccessible, unskippable intros, back button broken — and the web is better without it. But it dropped a powerful tool into the hands of people who weren’t “web programmers,” and a wave of weird, unexpected work came out the other side. Product configurators, drag-and-drop, loading screens as a creative medium. Some of it was bad. That was the point.
Today’s web is standardized, accessible, and banal. LLMs trained on that banality generate sites that are technically fine and all look the same. An April 2026 study found that by mid-2025 roughly 35% of newly published websites were AI-generated. The difference from Flash is that nobody is making creative decisions: replication, not experimentation.
The business argument follows from abundance:
- “We built an app” is no longer special
- Professional-looking design is not a differentiator
- Standard feature sets replicate instantly
- Even a genuinely novel output gets copied fast
What stays scarce: spotting problems everyone overlooked, reframing familiar things radically, inventing a new interaction or category, keeping unusual choices coherent, and continuing to evolve after you’re copied.
One idea isn’t a moat — the habit of generating ideas is. The author’s suggested use of AI is as an experiment multiplier, never a decision-maker: ask the model to list the assumptions behind your plan, pick one and imagine it being wrong, then have it critique your approach as someone who has to actually use the thing. Expect nonsense in the output; sorting what’s plausible from what isn’t is itself the exercise.
The line worth keeping: “A competitor cannot acquire all of that by prompting Claude.”