Solid trajectory from Kant’s stuck-in-a-simulation to Hegel situating the thing itself socio-historically. It was strange then that you circled back to Kant and found the solution to be the categorical imperative and the kingdom of ends. This seems like a step back for reasons Hegel suggested- Kant’s solution is sort of an empty signifier, a simulation in a simulation.
IMO the next step from Hegel is Nietzsche. He would agree that autonomy in itself is a dead end, but instead of going back to Kant (back to the simulation) he shifts the focus to living fully ie transcendence (think Emerson). This is in line with Taylor calls contact theory (in opposition to the simulation).
Living fully can still be socially situated like Hegel/Taylor says (he idolizes the Greeks for instance), but it avoids the last man stagnation that we get from plugging back into the simulation. And if you think Nietzsche is a bridge too far, then I’d suggest Huxley or Maslow for a gentler formulation.
I read Hegel as naturalizing the categorical imperative within the social practices of an ethical community, not as discarding the idea altogether. The issue with Kant's formulation is its discursive formality, which leads to a one-sided deformation of autonomy as consisting in the authority to will an ethical maxim without a co-relative responsibility to the preexisting attitudes of your normative community. Kant explicates the formal logic of the Golden Rule / "cooperate in the repeated prisoner's dilemma" equilibrium, but the formalism is empty without an antecedent social practice. The mere commandment that "Thou shalt not lie" has much less motivational force than belonging to a community in which a norm against lying is internalized. The idea I end on is that suitable a multi-agent learning environment may be necessary for getting advanced AI to bootstrap genuine autonomy, and if such agents include humans, to thereby induce the AI to see humans as "ends in themselves." There is no categorical imperative or "AI constitution" put in by hand. Cooperation needs to be emergent and retain the capacity to adapt and coevolve with human culture.
The issue there (which I think Nietzsche picks up on) is that getting socialized into a community isn't a stable solution if the community is itself off track. This is what I was getting at in my other comment with Taylor's critique of Reform and Habermas's critique of the system taking over the lifeworld.
It would be nice to get socialized into a community approximates Kant's kingdom of ends, or is on its way to doing so (per Hegel's idealism). In the same way it would be nice to have an AI training simulation that gets them in line with Kant's kingdom of ends (which is what you're saying). But I don't think it's safe to assume that we have that now or are on track for it (because of trends like Reform and the system).
In other words I think the society (or the AI system within it) needs to be oriented towards something like Taylor's fullness in order for Hegel/Kant's idealism to work, and I don't think we can take this presupposition for granted. Kant takes it for granted in the top-down way like you note, but Hegel too takes it for granted in the bottom-up way. It’s why Marx thought he had to change the world instead of interpreting it, and it’s why Taylor too says Hegel’s optimistic teleology is no longer valid.
But like I said, this is outside the scope of your argument so far. (Unless you have another essay on deck~)
"takes-it-for-granted" does this mean they throw it in the gap, or their gap, or the others' gap/s,— where there is a gap, that's where we judge everything, who to blame, who to credit.
These days I feel that's we are dancing, either inthe gap, or at least about the gap/s, pivoting from the logical hindsight of yesterday into hangover of that logic/determinism with whatever it is we dance/live/thunk/feel. It's like we cannot see how we see, but blame the shadows for the pain when we stub our toes. Either the shadow has agency and is a demon that is out to get us, or the shadow is some structure (emergent or top-down) that oppresses us.
The biggest issue with the acc message is that is has basically become: "you will die, and everything you care about, and there is nothing you can do about it, so you should welcome it."
Your work is the most exciting stuff in the AI space today. It makes me, a CS guy, want to study philosophy and liberal arts.
Thank you!
In my perspective, AI itself is a way for "reason" to emancipate from our evolutionary structures.
i am become ubermensch
good one
I wish AI could condense this into a lingo that I would have less to work on.^^
Solid trajectory from Kant’s stuck-in-a-simulation to Hegel situating the thing itself socio-historically. It was strange then that you circled back to Kant and found the solution to be the categorical imperative and the kingdom of ends. This seems like a step back for reasons Hegel suggested- Kant’s solution is sort of an empty signifier, a simulation in a simulation.
IMO the next step from Hegel is Nietzsche. He would agree that autonomy in itself is a dead end, but instead of going back to Kant (back to the simulation) he shifts the focus to living fully ie transcendence (think Emerson). This is in line with Taylor calls contact theory (in opposition to the simulation).
Living fully can still be socially situated like Hegel/Taylor says (he idolizes the Greeks for instance), but it avoids the last man stagnation that we get from plugging back into the simulation. And if you think Nietzsche is a bridge too far, then I’d suggest Huxley or Maslow for a gentler formulation.
I read Hegel as naturalizing the categorical imperative within the social practices of an ethical community, not as discarding the idea altogether. The issue with Kant's formulation is its discursive formality, which leads to a one-sided deformation of autonomy as consisting in the authority to will an ethical maxim without a co-relative responsibility to the preexisting attitudes of your normative community. Kant explicates the formal logic of the Golden Rule / "cooperate in the repeated prisoner's dilemma" equilibrium, but the formalism is empty without an antecedent social practice. The mere commandment that "Thou shalt not lie" has much less motivational force than belonging to a community in which a norm against lying is internalized. The idea I end on is that suitable a multi-agent learning environment may be necessary for getting advanced AI to bootstrap genuine autonomy, and if such agents include humans, to thereby induce the AI to see humans as "ends in themselves." There is no categorical imperative or "AI constitution" put in by hand. Cooperation needs to be emergent and retain the capacity to adapt and coevolve with human culture.
The issue there (which I think Nietzsche picks up on) is that getting socialized into a community isn't a stable solution if the community is itself off track. This is what I was getting at in my other comment with Taylor's critique of Reform and Habermas's critique of the system taking over the lifeworld.
It would be nice to get socialized into a community approximates Kant's kingdom of ends, or is on its way to doing so (per Hegel's idealism). In the same way it would be nice to have an AI training simulation that gets them in line with Kant's kingdom of ends (which is what you're saying). But I don't think it's safe to assume that we have that now or are on track for it (because of trends like Reform and the system).
In other words I think the society (or the AI system within it) needs to be oriented towards something like Taylor's fullness in order for Hegel/Kant's idealism to work, and I don't think we can take this presupposition for granted. Kant takes it for granted in the top-down way like you note, but Hegel too takes it for granted in the bottom-up way. It’s why Marx thought he had to change the world instead of interpreting it, and it’s why Taylor too says Hegel’s optimistic teleology is no longer valid.
But like I said, this is outside the scope of your argument so far. (Unless you have another essay on deck~)
"takes-it-for-granted" does this mean they throw it in the gap, or their gap, or the others' gap/s,— where there is a gap, that's where we judge everything, who to blame, who to credit.
These days I feel that's we are dancing, either inthe gap, or at least about the gap/s, pivoting from the logical hindsight of yesterday into hangover of that logic/determinism with whatever it is we dance/live/thunk/feel. It's like we cannot see how we see, but blame the shadows for the pain when we stub our toes. Either the shadow has agency and is a demon that is out to get us, or the shadow is some structure (emergent or top-down) that oppresses us.
The biggest issue with the acc message is that is has basically become: "you will die, and everything you care about, and there is nothing you can do about it, so you should welcome it."
You can see why reasonable people might object.