I don’t see anything that addresses the situation in which one company tries to take over the world using its AGI, or in which an AGI acting on its own initiative tries to take over the world, etc. Did I miss something?
Actually the OGI-1 model (and to a lesser extent, the OGI-N model) does do something important to address loss of control risks from AGI (or ASI): it reduces competitive race dynamics.
There are plausible scenarios where it is technically possible for a lab to safely develop AGI, but where doing so would require them to slow down development. When they are competitively racing against other AGI projects, the incentives are (potentially much) stronger to proceed with risky development. But when a lab doesn’t have to worry about competitors, then they at least have an opportunity to pursue costly safety measures without sacrificing their lead.
Actually the OGI-1 model (and to a lesser extent, the OGI-N model) does do something important to address loss of control risks from AGI (or ASI): it reduces competitive race dynamics.
There are plausible scenarios where it is technically possible for a lab to safely develop AGI, but where doing so would require them to slow down development. When they are competitively racing against other AGI projects, the incentives are (potentially much) stronger to proceed with risky development. But when a lab doesn’t have to worry about competitors, then they at least have an opportunity to pursue costly safety measures without sacrificing their lead.