AI will never be able to replace a human board member. I think that's what bothers technology people the most here. It's quite asymmetric. At best, LLM agents will be granted certain non voting advisory roles, but anything the llm generates will need a human to assume responsibility.

The point of a human being on the board is that they can be made to suffer the consequences of bad decisions. Executive pay might seem astronomical, but it is often commensurate with the level of stress and responsibility involved. It's easy to look at the perks and not see the struggle behind them. No one in media is going to get a lot of clicks if they publish articles about how being a CEO is actually really hard and maybe some of what they're doing is actually justified once you assume all of the same context and stress they have.

If we want to do to the executive staff what is being done to the technical staff, I'm afraid we will need to first figure out a way to make the AI experience human emotions as strongly as we do. It often sounds fun in first order terms to threaten to fire a CEO and replace them with a clanker, but have we considered the consequences of this? What would it be like to be under the management of an emotionless robot? Being managed by a robot vs managing robots are wildly different things to me.