I note up front that at least Daniel Kokotajlo has indeed adjusted his estimates, and has moved his median from ‘AI 2027’ to ‘AI 2028’ based on events since publication
This is not quite right. My timelines move dup to median 2028 before we published AI 2027 actually, based on a variety of factors including iteratively updating our models. But it was too late to rewrite the whole thing to happen a year later, so we just published it anyway. I tweeted about this a while ago iirc.
First footnote, right on front page: “We disagree somewhat amongst ourselves about AI timelines; our median AGI arrival date is somewhat longer than what this scenario depicts. This scenario depicts something like our mode. See our timelines forecast for more details.”
...But no, I guess it doesn’t say that Daniel K’s median is 2028, iirc. (haven’t checked, going on memory) The timelines forecast has Eli and Nikola’s medians because they were the ones who wrote it.
This is not quite right. My timelines move dup to median 2028 before we published AI 2027 actually, based on a variety of factors including iteratively updating our models. But it was too late to rewrite the whole thing to happen a year later, so we just published it anyway. I tweeted about this a while ago iirc.
Does it say this somewhere on the website?
First footnote, right on front page: “We disagree somewhat amongst ourselves about AI timelines; our median AGI arrival date is somewhat longer than what this scenario depicts. This scenario depicts something like our mode. See our timelines forecast for more details.”
...But no, I guess it doesn’t say that Daniel K’s median is 2028, iirc. (haven’t checked, going on memory) The timelines forecast has Eli and Nikola’s medians because they were the ones who wrote it.