GGP comment seemed to mostly be caring about personal computers, in which field we do have most the computing power we will need.
For those kinds of hard problems, the advances seem to be coming in massively parallel computing of the kind that is advancing (the GPGPU has transformed ML, for example).
For those kinds of hard problems, the advances seem to be coming in massively parallel computing of the kind that is advancing (the GPGPU has transformed ML, for example).