The assumption is that AI builds momentum in nuclear power to the point where it can be used elsewhere. I've kept hearing that nuclear is very expensive mainly due to lack of scale.
The reality is that human energy consumption is expanding faster than nuclear is likely to be built out.
It's a tall ask for nuclear to cover projected additional demand (see: global energy consumption data over time), it's even more of an ask to expect it to make inroads on existing demand also.
Nuclear is also expensive in comparison to alternatives (depending upon local expertise): Australia's costings (cold start on nuclear power tech) are that the better ROI over next four decades on energy investment now is not nuclear (2024 CSIRO energy report).
defrost|1 year ago
It's a tall ask for nuclear to cover projected additional demand (see: global energy consumption data over time), it's even more of an ask to expect it to make inroads on existing demand also.
The closest thing we have for examples of nuclear at scale is Korea in recent past decades and currently China right now: https://itif.org/publications/2024/06/17/how-innovative-is-c...
Nuclear is also expensive in comparison to alternatives (depending upon local expertise): Australia's costings (cold start on nuclear power tech) are that the better ROI over next four decades on energy investment now is not nuclear (2024 CSIRO energy report).