AI compresses the development cycle. The interesting question is what you do with the room it gives you.
Here’s the thesis the whole publication runs on: used well, these tools shrink the time and cost of a development cycle. That’s it. That’s the prize. Everything else is detail.
But “cheaper and faster” undersells what actually happens, because the savings don’t have to come out as savings. When AI compresses the cycle, you get a budget — of time, money, and engineering attention — and you get to decide where it goes.
Option one: you spend it. You ship the same product for less, in less time. Lower cost, faster to market, real margin. For a lot of programs, that’s exactly right.
Option two: you reinvest it. You take the time and money AI gave back and pour it into quality — more validation, better tolerancing, a more robust supply chain, the design review you always skip when the schedule gets tight. Same time, same budget as before, but a materially better product out the door.
Either way you come out ahead. Either you get the same product cheaper and faster, or a better product for the same cost and time. There’s no version of this where doing it well leaves you worse off. The development cycle is the constraint that governs everything in hardware — what you can afford to build, how good it can be, how fast it reaches the people who need it. Anything that loosens that constraint is worth taking seriously.
That’s why we’re doing this. Not because AI is exciting. Because superior products, shipped at a lower cost of money and time, is the entire game — and for the first time, the tools to get there are real enough to be worth an honest look.
First, an honest look.