I put it at about 60% likely that AI will solve fusion as a power source for us.

If that happens, even if we spend $100 trillion dollars building out AI datacenters it will have been worth it.