How far can we push computing?
Feynman stood up in 1959 and basically dared engineers to write the entire Encyclopedia Britannica on the head of a pin. We did that and then some, and now the pin is the size of a few atoms and the electrons are getting weird about it. The part I keep circling back to is the energy cost of erasing information. There's a hard floor down there, set by thermodynamics, and we're close enough to it that the old trick of just shrinking everything no longer works.
I'd want to know what happens when we stop treating computation as something that has to happen in silicon at all. Feynman himself floated the idea of quantum simulation, and that thread has pulled us into a whole different way of thinking about what a calculation even is. Maybe the next leap isn't a faster chip but a different definition of the problem.
I'm not sure we'll ever hit an absolute wall. Every time we think we've found the bottom, someone figures out a cleverer way to rearrange the atoms. But it's a strange feeling to watch an industry that ran on smaller and faster start talking about information as a physical substance with real costs. That's the part that feels like a genuine shift.