Interesting to see upvotes on this post now, five years later, and return to it.
I think my estimate of impact was too simplistic and probably erred in the optimistic direction. That said, I don’t think it’s obvious that MIRI moving to Singapore in 2020 would have been a bad idea. Consider how UKAISI seems to have grown super fast in the two years, due to having an awesome core of people to seed it & initial government support. I think it’s likely that if MIRI had gone to Singapore in 2020, by 2023 there would have been similar government support + talent available and Singapore would have their own equivalent of UKAISI. So, it comes down to the opportunity cost of MIRI doing that vs. doing what they in fact did.
Interesting to see upvotes on this post now, five years later, and return to it.
I think my estimate of impact was too simplistic and probably erred in the optimistic direction. That said, I don’t think it’s obvious that MIRI moving to Singapore in 2020 would have been a bad idea. Consider how UKAISI seems to have grown super fast in the two years, due to having an awesome core of people to seed it & initial government support. I think it’s likely that if MIRI had gone to Singapore in 2020, by 2023 there would have been similar government support + talent available and Singapore would have their own equivalent of UKAISI. So, it comes down to the opportunity cost of MIRI doing that vs. doing what they in fact did.