Yeah, I don't think 1 THz is achievable, but we can likely do substantially better than 4 Ghz, even when we can't go smaller. Then there are also gains to be made from building bigger, building in 3 dimensions, decreasing waste heat, etc. I suspect that innovation with chips will get bumpier, but likely overall maintain the pace of innovation for a couple decades more - which is all we need to get into really interesting territory re AI and other things.