Two other interesting tidbits to contextualize the heavy investments from AI/Cloud hyper scalers into Nuclear
1) Based on my contacts, I hear that nuclear simulations can be a very lucrative avenue for AI Builders. More nuclear would increase the demand for the hyper scalers services.
2) Similarly this might have some good synergy with the HPC and super computing plays by companies. I know Google is making movies in the HPC market, and I'm sure others have their own. As with AI/cloud, I see some good cross selling potential.
These: “It’s an engineering problem. It’s a construction problem. It’s an economic problem” are not the problems. Ok, they are the proximate issues, but they are only relevant if you presume nuclear is a good idea.
The problem with nuclear isn’t engineering, construction, or economics, it is time. Whatever size you build, the reactor and its fuel will be deadly for millennia. Moreover, even in the short-term (100 years or so), a reactor requires very diligent management, regulation and oversight.
There is no guarantee that our democracy will last the year, and no civilization has operated continuously for more than a few hundred. The fact of Roman ruins should demonstrate that.
Humans are not capable of the level of discipline and societal stability to maintain reactors and protect the earth from radioactivity. As such, we do not have the right to create more reactors.
To do otherwise is to condemn the future for a greedy present.
We have a little different POV. Fractal technology eliminates the need for most data centers as we are proving in the electric utility industry - at least we cannot find any application, even AI that requires a data center. We think there are new tech stacks, eliminating I/O that will solve part of the problem. Jay@FractalWeb.App
Two other interesting tidbits to contextualize the heavy investments from AI/Cloud hyper scalers into Nuclear
1) Based on my contacts, I hear that nuclear simulations can be a very lucrative avenue for AI Builders. More nuclear would increase the demand for the hyper scalers services.
2) Similarly this might have some good synergy with the HPC and super computing plays by companies. I know Google is making movies in the HPC market, and I'm sure others have their own. As with AI/cloud, I see some good cross selling potential.
Very interesting, thanks for sharing!
These: “It’s an engineering problem. It’s a construction problem. It’s an economic problem” are not the problems. Ok, they are the proximate issues, but they are only relevant if you presume nuclear is a good idea.
The problem with nuclear isn’t engineering, construction, or economics, it is time. Whatever size you build, the reactor and its fuel will be deadly for millennia. Moreover, even in the short-term (100 years or so), a reactor requires very diligent management, regulation and oversight.
There is no guarantee that our democracy will last the year, and no civilization has operated continuously for more than a few hundred. The fact of Roman ruins should demonstrate that.
Humans are not capable of the level of discipline and societal stability to maintain reactors and protect the earth from radioactivity. As such, we do not have the right to create more reactors.
To do otherwise is to condemn the future for a greedy present.
We have a little different POV. Fractal technology eliminates the need for most data centers as we are proving in the electric utility industry - at least we cannot find any application, even AI that requires a data center. We think there are new tech stacks, eliminating I/O that will solve part of the problem. Jay@FractalWeb.App