Umm… I have already warned that inner troubles of the American administration might cost OpenBrain lots of compute, which can influence the AI race.
I have also made a comment where I tried to show that the US leadership would be undermined by the Taiwan invasion unless the US domestic chip production dominates the Chinese one. It would be especially terrifying to discover that both OpenBrain and DeepCent have similar amounts of compute while neither side[1] can increase these amounts faster than the other (and I did imply something similar in the comment I made!), since neither the US nor China can slow down without an international deal. And a hypothetical decay of the US makes the matters worse for the US.
Moreover, I have mentioned the possibility that the US administration realises that they have a misaligned AI, but without the AI-driven transformation of the economy the US will be unable to produce more chips and/or energy, leaving China with leadership. Then the US could be forced to let the AI transform the economy. Or threaten to unleash the misaligned AI unless China somehow surrenders its potential leadership...
Could you ask the AI-2027 team to reconsider the compute forecast and estimate the influence of the revised compute and power of the AIs on the other aspects of the scenario?
The slowdown ending of AI-2027.com had OpenBrain receive compute by merging with its rivals. The collapsed section about the Indo-Pakistan nuclear war (which would be equivalent to the Taiwan invasion in 2025) in my comment describes a situation where OpenBrain and its former rivals have done a number of computations similar to DeepCent and its former rivals.
Umm… I have already warned that inner troubles of the American administration might cost OpenBrain lots of compute, which can influence the AI race.
I have also made a comment where I tried to show that the US leadership would be undermined by the Taiwan invasion unless the US domestic chip production dominates the Chinese one. It would be especially terrifying to discover that both OpenBrain and DeepCent have similar amounts of compute while neither side[1] can increase these amounts faster than the other (and I did imply something similar in the comment I made!), since neither the US nor China can slow down without an international deal. And a hypothetical decay of the US makes the matters worse for the US.
Moreover, I have mentioned the possibility that the US administration realises that they have a misaligned AI, but without the AI-driven transformation of the economy the US will be unable to produce more chips and/or energy, leaving China with leadership. Then the US could be forced to let the AI transform the economy. Or threaten to unleash the misaligned AI unless China somehow surrenders its potential leadership...
Could you ask the AI-2027 team to reconsider the compute forecast and estimate the influence of the revised compute and power of the AIs on the other aspects of the scenario?
The slowdown ending of AI-2027.com had OpenBrain receive compute by merging with its rivals. The collapsed section about the Indo-Pakistan nuclear war (which would be equivalent to the Taiwan invasion in 2025) in my comment describes a situation where OpenBrain and its former rivals have done a number of computations similar to DeepCent and its former rivals.