You say your estimate for when alignment will be solved is from a “Gaussian distribution… used for illustration purposes only”. How do you intend the graphs/numbers based on this estimate to be interpreted?
It is the naive estimate and we would need a broader range of others’ estimates (e.g. on Metaculus) before we make concise predictions based on this. I also wouldn’t expect it to follow a gaussian at the moment but some mixture distribution of several parameters.
You say your estimate for when alignment will be solved is from a “Gaussian distribution… used for illustration purposes only”. How do you intend the graphs/numbers based on this estimate to be interpreted?
It is the naive estimate and we would need a broader range of others’ estimates (e.g. on Metaculus) before we make concise predictions based on this. I also wouldn’t expect it to follow a gaussian at the moment but some mixture distribution of several parameters.