I agree with this picture pretty closely. I think 2025 is unlikely, because there are probably a couple of tricks left to invent. But 2026-30 carries a lot of probability mass, and the rest is spread out over decades.
I agree with this picture pretty closely. I think 2025 is unlikely, because there are probably a couple of tricks left to invent. But 2026-30 carries a lot of probability mass, and the rest is spread out over decades.