This of course depends on the time frame one wants to achieve, but assuming we want to remove carbon dioxide as quickly as we added it, that means we would need some operation comparable in size to the global coal, oil and gas industry extracting carbon dioxide from the atmosphere around the clock. And while I agree with you, exponential growth is in some sense hard to predict, I don't really see us starting such a gigantic endeavor unless we are absolutely forced to by the consequences of climate change.

You make solar power cheaper than coal, and atmospheric carbon useful for something, and you'll soon have to regulate it extraction, not emission.

Today's computers were perfectly obvious to Gordon Moore in his 1965 paper. By contrast nobody today is writing papers about the inevitable expansion of carbon-free energy production.

Some folks are certainly hinting at it. http://rameznaam.com/2015/08/10/how-cheap-can-solar-get-very...

Various forms of next generation nuclear would meet that requirement right? People are certainly writing papers on building better nuclear power plants. Doesn't even have to be something exotic like fusion.

If there is some kind of nuclear power that is millions of times cheaper per unit output than it was in 1965, I am unaware of it.