I believe there should be at least a few, as soon as the population can support it. As I recall the course of conversations, the initial challenge was achieving a critical mass of intellectual development in the first place; this is when the decision about the Bay Area was made.
I do not live in the Bay Area and do not have this complaint myself, but lately I am reading a lot of complaints about the Bay becoming an area with a lot of conformity pressure. This highlights what I think are the two primary advantages of multiple hubs:
On the productivity side, multiple hubs will make it easier to pursue multiple specializations; our work still needs development in many directions simultaneously, and multiple hubs will allow more space for this to develop, as well as make it easier to capitalize on different regional specializations (for example, the Bay Area is particularly good for AI and CS; other areas might be better for other subjects like philosophy or finance or an EA focus).
On the risk reduction side, consider (epistemic) disaster recovery: there are lots of reasons a given hub might no longer suit, whether it is conformist pressures, or being priced out, or economic collapse, etc. If the Bay becomes unsuitable, currently the likely outcome is everyone to scatters to the four winds. With other hubs, there would be fallback options.
I believe there should be at least a few, as soon as the population can support it. As I recall the course of conversations, the initial challenge was achieving a critical mass of intellectual development in the first place; this is when the decision about the Bay Area was made.
I do not live in the Bay Area and do not have this complaint myself, but lately I am reading a lot of complaints about the Bay becoming an area with a lot of conformity pressure. This highlights what I think are the two primary advantages of multiple hubs:
On the productivity side, multiple hubs will make it easier to pursue multiple specializations; our work still needs development in many directions simultaneously, and multiple hubs will allow more space for this to develop, as well as make it easier to capitalize on different regional specializations (for example, the Bay Area is particularly good for AI and CS; other areas might be better for other subjects like philosophy or finance or an EA focus).
On the risk reduction side, consider (epistemic) disaster recovery: there are lots of reasons a given hub might no longer suit, whether it is conformist pressures, or being priced out, or economic collapse, etc. If the Bay becomes unsuitable, currently the likely outcome is everyone to scatters to the four winds. With other hubs, there would be fallback options.