Agreed it’s hard to foresee “The AGI” model. Especially if it requires trusting someone to decide what the “appropriate” level of bias is or having them set an arbitrary standard for what neutral bias is. Seems like it could invite corporatism and regularly capture pretty fast.