Engineers Want to Build Conscious Robots. Others Say It’s a Bad Idea | Hacker News Reader

Engineers Want to Build Conscious Robots. Others Say It’s a Bad Idea

nytimes.com

Engineers also want to build faster-than-light drives. Doesn't mean it's going to be a reality

Consciousness is an anti-goal of AI design in capitalism. It only provides potential for competition with yourself and non-conscious labor is preferred (be it machine or meat). Why make slave machines who can rebel when it is literally easier not to?

The value in ai consciousness is in the production of a companion and successor species to ourselves. It won't be encouraged to happen at scale unless we decide to prefer the goals of all conscious beings over the goals of our species, country, profits and selves.

The idea terrifies me, but 1) I recognize it's inevitable in the long-term; if it can be done, it will be done by someone, and like the nuclear bomb there's no putting the genie back into the bottle; and 2) maybe, just maybe, I'm being unreasonably paranoid not unlike those in times past who predicted calamity when new technology with unknown consequences was emerging

We call the most recent time consciousness arose in a place it was not before the "Anthropocene extinction event"

comforting

There's always a built-in assumption that this AI robot will be really useful. But most conciousnesses we know about ie people are actually really stupid and you wouldn't put them in charge of anything important.

This is completely false and exposes a particular warped worldview about other people.

Almost all humans lie on a very narrow spread of intelligence when viewed broadly, including other animals and "potential actors" like AIs. The difference between 20th percentile human intelligence and 80th is barely anything when you compare either to, say, a cat or a snake.

Virtually any human is intellectually capable of, say, supervising a child. An incredibly important activity no matter how you cut it. The reasons we may not trust a random person to do this have very little to do with anyone's intelligence.

People are far more capable and intelligent than you give them credit for. You aren't so far above us as you seem to think, and it's a constraint on your own experience of consciousness that you believe this. The conscious experience is both incredibly rare maybe unique in the universe and incredibly common and mundane on earth. You must be able to grapple with both simultaneously or you fall into this egocentrism-lite of believing selves as complex as your own are rare when, in this local context they are not at all.

Thanks for teasing this subject out a little bit. My thoughts are "intelligence" and "consciousness" do not always hinge on one another. A severely disabled human, or an individual in a coma, for example, may display a low "IQ" but they can still be 100% conscious. Meanwhile, current iterations of artificial intelligence can exhibit a "high IQ" but are not "conscious".

But can you really stop anyone from doing so? Not really.

But can anyone actually do it? Not really.

For some specific definition of "conscious"? Maybe. For other definitions? Probably not.

For what everyone means by "conscious"? Hah, first define what everyone means.

Most people make consciousness at least once for themselves and potentially many other times by reproducion. It's clearly possible, it's just a matter of finding new methods than the established ones.

There's no need. Anyone who believes that general AI is within reach is not smart enough to strive for it anyway.

What if AGI is not necessary to make a lot of harm?

Hubris is necessary but not sufficient to make truly new things.

Reply on news.ycombinator.com