That's a sharp point about the cost curve not being a decisive boundary. You're right, one big breakthrough, like a new engine design for a car, could entirely shift the economic viability of getting to the next level of intelligence. The current cost structure could snap instantly.