We will need MILLIONS of people in the field.

it's a real issue but I'm not sure it's the mass mobilization problem the article makes it out to be. pretty sure it's growing extremely fast, and it's a new field in general. the models will keep growing in capability because of human game theory (I need to make my model better because my competitors are). it's not practical to expect millions in headcount to move from frontier capability research to safety