🇳🇱
So I started a company and received funding to build the next generation of #AI / #agi
I know there are a lot of fears around AI, and I share in most of them. As such a top priority will be for me to address the ethical considerations. I am still branstorming how that should look but I want it to be an open forum where everyone can contribute to solve the ethical considerations.
For now I'd love to hear input from people on how one could build a community to address and solve ethical concerns in AI / AGI
@freemo how do we build ethical humans?
Or protect us against the unethical ones.
In the case of AGI maybe its ok if we build some that are unethical as long as we build a system where the ethical AIs can keep the unethical ones in check.
> is being ethical something that can be quantified and measured?
In theory yes, in practice, sorta.
Being ethical is simply the act of reducing suffering and maximizing happiness. Since these are just chemical states in ones mind it could, in theory, be quantified if you could scan everyone brain effected. In practice we cant do that so we have to observe people to infer their happiness. But yes it is objective in that sense.
The bigger issue is moral intent vs moral action, both have their problems... Moral intent doesnt matter here much if it doesnt produce moral action, so we dont even need to think much on that as it isnt relevant here. Moral action on the other hand could be qualntified as said easrlier but even if we could scan brains part of the problem is you'd have to quantify the long term effect and compare to what that effect would be if you did nothing. Since we dont have time machines we cant provide an actual comparison anyway.
That said subjectively it can be infered.
> I'm thinking of Rice's Theorem which many people in the AGI community like to discount but no one has ever convinced me about that.
Rice's theorem is valid.
> Merry Christmas, I haven't even started drinking yet.
Merry Christmas to you too, have a wonderful one!
@freemo is being ethical something that can be quantified and measured?
I'm thinking of Rice's Theorem which many people in the AGI community like to discount but no one has ever convinced me about that.
Merry Christmas, I haven't even started drinking yet.