People were seriously asking back in 2011: when will we be able to program on a phone? The most optimistic estimates then would have sounded aggressive set in 2016. It’s 2026. Still not here.
Why? Performance stopped being the issue long ago, and the wall isn’t made of bricks. The phone simply never gave you root.
iOS’s sandbox won’t let you run a compiler freely: switch to your browser to check the docs, and it kills your build process. Android is looser, but Termux-style solutions remain hobbyist toys, never a mainstream path. Fifteen years, chip performance multiplied beyond counting, and the wall hasn’t moved an inch.
Today that wall doesn’t just block compilers. It blocks agents. An agent acts through permissions: reading your files, tapping your screen, running your scripts. Without permission, AI has no hands. A handless agent, however smart, is a consultant — consultants only advise; agents act.
Which makes some product choices curious. Everyone proclaims that in the AI era everyone is a developer — and then tells you that to implement one feature, you must go through the developer platform. That itself is pretty funny.
Where’s the flaw? Designing for users assumed to be stupid. Presuming you understand nothing and must follow its flow; paving every road for you and sealing every fork.
That said, not everyone deserves the tar brush. Anthropic’s product opens to a single box, no buttons — if the user is dumb, that box leaves them helpless. But that’s precisely respect: we assume you can express yourself. An empty box is a mirror; what it reflects is whether your own need is clear.
So the dividing line isn’t the box. It’s the permission. Behind the box, do you get the whole machine? Android at least keeps a developer mode; some systems seal off even that fantasy. The subtext is identical: users don’t deserve full control of their own machine.
Permission is the AI’s hand. Tie the hand, and no amount of intelligence makes it more than a talking consultant.
