The probability of achieving AGI within ten years is 95% under extreme hypothetical scenarios (e.g., global disruptions like war or supply chain collapse), but confidence in verifiable tasks is very high.
Maybe the irreducible uncertainty puts us at 95%, where you get to things like multiple companies having internal turmoil, Taiwan gets invaded, all the fabs get blown up by missiles. Now you've jinxed us, Dario. You could construct a 5% world where things get delayed for ten years. There's another 5% which is that I'm very confident on tasks that can be verified.