Robin thinks emulations will probably come before AI, that non-emulation AI would probably be developed by large commercial or military organizations, that AI capacity would ramp up relatively slowly, and that extensive safety measures will likely prevent organizations from losing control of their AIs. He says that still leaves enough of an existential risk to be worth working on, but I don't know his current estimate. Also, some might differ from Robin in valuing a Darwinian/burning the cosmic commons outcome.

I don't know of any charitable contributions Robin has made to any organization, or any public analysis or ranking of charities by him.