Back when many expected takeoff in 2027 or so, it was pretty reasonable to assume that the probability of a conflict entirely unrelated to AI was low.[3] But the forecasters behind AI 2027 now expect takeoff in the 2030s.
I’m saddened that this is the takeaway from our new model! It seems misleading. Here is a graph of my+Eli’s timelines over time:
I’m sorry if I misrepresented your beliefs! I was mainly basing this on the timeline in the new model (using Eli’s median parameters), in which Automated Coder (AC) is reached in 2031 and ASI in 2034, which I thought was well-described by “takeoff in the 2030s”. Does AGI in this plot refer to the same thing as ASI in the model? If your timelines have ASI in 2030 (and presumably AC somewhat earlier), then I can revise what I wrote to try to reflect that.
I guess another issue is that even if the median is in the early 2030s, that still leaves half the probability mass earlier than that, so what I wrote might be misleading if it’s taken to imply that you definitely don’t expect takeoff before 2030.
I doubt that the potential misconception is actually relevant to the issue of modifying the model due to Taiwan being invaded. Yourtwoforecasts imply that P(automated coder by Jan 2028) is at most 27% and Eli’sforecasts have P(automated coder by Jan 2028) is less than 25%. Similarly, neither you nor Eli believe that P(superhuman coders appear by Jan 2029) is at least 40%. Were Taiwan to be invaded during Jan 2027 and extrapolation of pre-invasion trends with exponential growth of compute to yield automated coders after Jan 2029, OpenBrain would no longer have the way to reach the SC level by increasing the compute tenfold. Instead I would expect the USG to merge the companies, have OpenBrain reach Agent-4 far later, try to slow down and realize that they can’t gather any more resources. How would DeepCent’s compute increase after the invasion?
I’m saddened that this is the takeaway from our new model! It seems misleading. Here is a graph of my+Eli’s timelines over time:
https://x.com/DKokotajlo/status/1992316620254155028
I’m sorry if I misrepresented your beliefs! I was mainly basing this on the timeline in the new model (using Eli’s median parameters), in which Automated Coder (AC) is reached in 2031 and ASI in 2034, which I thought was well-described by “takeoff in the 2030s”. Does AGI in this plot refer to the same thing as ASI in the model? If your timelines have ASI in 2030 (and presumably AC somewhat earlier), then I can revise what I wrote to try to reflect that.
I guess another issue is that even if the median is in the early 2030s, that still leaves half the probability mass earlier than that, so what I wrote might be misleading if it’s taken to imply that you definitely don’t expect takeoff before 2030.
I doubt that the potential misconception is actually relevant to the issue of modifying the model due to Taiwan being invaded. Your two forecasts imply that P(automated coder by Jan 2028) is at most 27% and Eli’s forecasts have P(automated coder by Jan 2028) is less than 25%. Similarly, neither you nor Eli believe that P(superhuman coders appear by Jan 2029) is at least 40%. Were Taiwan to be invaded during Jan 2027 and extrapolation of pre-invasion trends with exponential growth of compute to yield automated coders after Jan 2029, OpenBrain would no longer have the way to reach the SC level by increasing the compute tenfold. Instead I would expect the USG to merge the companies, have OpenBrain reach Agent-4 far later, try to slow down and realize that they can’t gather any more resources. How would DeepCent’s compute increase after the invasion?