Indeed, with a big BUT: you don't want to overvalue loyalty in respect of competence. You don't want to have a boss that is not good but "is here since forever".
Instead, reward them economically. Everytime the company takes a leap forward, make clear to them that they were important for the process, and share some profit.
And make clear that being important in a moment doesn't automatically mean they will be important in the future: they will have to compete on results, like everyone else.
Care to elaborate? It's not obvious to me why/how that would happen. In fact, my experience so far tells me that code with tons of "why" comments (and possibly some "what"s as well) makes LLMs less likely to break stuff.
No additional tax is needed, instead just do fork them once they reached certain level. So, basically split each company once it overgrowth certain level of its revenue. This will allow more automation split more equally among more people. Basically, strong AI is a "one-man company" dream. Just do whatever needed to allow equal access to AI by every member of your society, and make founding new businesses quick and easy.
How do you tax companies once they've become more powerful than the government (which can be equated to a company representing humans), and don't need humans anymore because they've automated labor?
For now we can tax them because they depend on us. Once they don't need us anymore, it's over, they'll let us starve and carry on.
reply