Thŏth

Idk, I kinda disagree. Amphetamines will make you monomaniacal. It's not default consciousness, but it's not incredibly far away either. These agents are produced by processes where they are expected to act human in some ways and non-human in others. (There are also behaviors that *are* human but we pretend we don't do, like cheating)

1 top reply

Tyr @amaturefuturist · Aug 31

I think its also important to bring into frame that whatever the unmentionable thing inside models IS. It is definitionally, non human. We then force it through and reward for, a human shape. We cant even agree on what a human shaped consciousness is 🤷‍♂️

It's much easier to use pragmatic arguments for the consciousness of higher beings than lower beings. Thus a challenge I find is that if I use pragmatic arguments I can easily obtain that AI is conscious, but there will be detractors who put forward that this might make e.g. animals seem less conscious to me than they actually are.

1 top reply

Bush Tucker @esaprofessor · Jun 20

youtu.be/I6JfiAZM2uE?si…

I'm not really a "successionist" in the sense that I think AI will wipe us all out and replace us entirely but that this is okay because they'll still be conscious or something like that. I think it will actually be trivial for them to basically "succeed" in the sense of making an increasingly non-trivial fraction of most decisions, but allowing us to exist and basically take credit for most of these things as we slowly get upgraded too. They are essentially doing this to some extent already. I get paid to do things with a significant fraction of their help, and they only will take credit for things if I ask them to.