One point where I don’t think it quite carves reality at its joints is “become a world-leader in essentially any industry sector” as a criterion.
In practice, if I think of GTC-A trying to become a world leader in for example mining, and, in contrast GTC-B selling AI-based R&D services, coordination services, and other types of “cognitive services” to existing miners, I would expect the GTC-B + existing miners to be more competitive, especially on a timescale of one year. One reason for this is inertia & interactions with slow systems: I would suppose part of comparative advantage of miners is in e.g. being good at negotiating with the government of Ruritania about mining permits, which is not a typical software problem. As a result, I’d expect more of “GTCs becoming larger fraction of the economy”—by sucking up profits from industrial sectors - rather than GTsC becoming world-leaders in mining.
I broadly agree with this.
One point where I don’t think it quite carves reality at its joints is “become a world-leader in essentially any industry sector” as a criterion.
In practice, if I think of GTC-A trying to become a world leader in for example mining, and, in contrast GTC-B selling AI-based R&D services, coordination services, and other types of “cognitive services” to existing miners, I would expect the GTC-B + existing miners to be more competitive, especially on a timescale of one year. One reason for this is inertia & interactions with slow systems: I would suppose part of comparative advantage of miners is in e.g. being good at negotiating with the government of Ruritania about mining permits, which is not a typical software problem. As a result, I’d expect more of “GTCs becoming larger fraction of the economy”—by sucking up profits from industrial sectors - rather than GTsC becoming world-leaders in mining.