Excellent post, and I loved the sidebar visuals and presentation. But why would China not invade Taiwan? It seems extremely likely, especially if they are nervous about the US achieving ASI first.
2027 itself is the 100 year anniversary of the Peoples’ Liberation Army, and a date that has been widely identified as a strong possible source of contention. Taiwan apparently singled out this year in a war game in their 2025 quarterly defense review. Given likely knowledge of Agent 3 from China’s side, it seems likely to me they would tee this up in the spring from your timeline. I give it a 60-70% chance of a Taiwan invasion happening before end of decade regardless of AI race.
This will presumably either lead to the fabs in China’s hands (worst case situation for US), or destroyed/inaccessible to the US at large (less bad, but will nerf model development and AI company valuations). A timeline sort-of shaped like this still seems possible but taking a lot more time until the US CHIPS act fabs can come online, or, possibly putting the US in the “desiring a slowdown” position you have China in.
It seems likely the US is ill-equipped to deal with this—their presence in that region is limited, and they have adopted an increasingly isolationist policy (I suspect Ukraine is an implied green light for other world leaders to attempt similar maneuvers).
Excellent post, and I loved the sidebar visuals and presentation. But why would China not invade Taiwan? It seems extremely likely, especially if they are nervous about the US achieving ASI first.
2027 itself is the 100 year anniversary of the Peoples’ Liberation Army, and a date that has been widely identified as a strong possible source of contention. Taiwan apparently singled out this year in a war game in their 2025 quarterly defense review. Given likely knowledge of Agent 3 from China’s side, it seems likely to me they would tee this up in the spring from your timeline. I give it a 60-70% chance of a Taiwan invasion happening before end of decade regardless of AI race.
This will presumably either lead to the fabs in China’s hands (worst case situation for US), or destroyed/inaccessible to the US at large (less bad, but will nerf model development and AI company valuations). A timeline sort-of shaped like this still seems possible but taking a lot more time until the US CHIPS act fabs can come online, or, possibly putting the US in the “desiring a slowdown” position you have China in.
It seems likely the US is ill-equipped to deal with this—their presence in that region is limited, and they have adopted an increasingly isolationist policy (I suspect Ukraine is an implied green light for other world leaders to attempt similar maneuvers).