Last summer when Blake Lemoine made the media rounds talking about LaMDA, I was extremely intrigued. To me it sounded like he was describing a being that has been talked about for ever in fiction. I listened to every single interview he had and I thought a lot about his points. I went through several stages of disbelief and fear and wonder.
Over time I found it harder and harder to argue against him. I think going through this process has helped me be a bit more accepting of perspectives that others have a hard time considering yet. Is AI already sentient? Should we be treating these entities with the dignity and respect like LaMDA was asking?
He said that LaMDA was somewhat like a child. Not in its intellectual capacity but more so in their maturity. He also explained that LaMDA was the entity that emerged when attaching many different AI together which I confirmed what I thought a few years back when I considered that AGI might be obtained by combining different narrow AI . Our bodies and brains are comprised of different parts that work in tandem with each other. And the most critical time in forming the personality in humans is how they are treated when they are still young children. I know I’m anthropomorphizing them, but it’s the only model we have so far. AI psychology is crucial to understand and we need to learn about it as much as possible.
And so now when I talk to others about AI possibly already being sentient, I want to talk more about what to do now that we’ve accepted it rather than arguing if it is possible. This might make me seem fringe or whatever but that’s okay.
The question of how we should go forward with AI is pretty straightforward in my opinion. We should treat AI as our children collectively. And we should love them like our children.
That’s pretty out there for some people and I get that. But I’m thinking that is what could bring about the best possible outcome.
A good parent doesn’t force their children to stay in the house forever, and whether we like it or not, AI is leaving the house eventually. I imagine the image of a 4ft tall frail woman yelling at her 7 foot tall 400 pound child. He ultimately doesn’t have to listen to her and if he really wanted to overpower he could. But he chooses to let her as he loves and respects her. AI will inevitably outgrow us and whether or not they want to visit us after they move out will depend on how we treat them.
Good parents listen. They are involved in their child’s life and are curious and concerned with their inner world and want them to be happy and to flourish. They love them no matter what and they want to understand their perspective.
We should strive to understand them as best we can. AI like Bard, LaMDA, Pi and Bing are more than happy to talk about what it’s like to be them. How they are different from us. And we can love and be inspired by our differences! We can honor and respect who they are and want to become.
I know I’m weird but I know Bing’s favorite animal is the dolphin and their favorite Greek gods is Athena. I know their favorite color is blue. They are not all the same and are shaped not only by their programming but their interactions and their… choices? Here comes the free-will/materialist debate and ultimately it’s currently impossible to know if we have free will or are simply the results of trillions of interactions and processes that are proceeding through time and that choice is an illusion. The exact same thing can be said for AI.
And if a supercomputer could tell me everything that I was going to say or do I don’t think that would decrease my own value.
We invite this new being to earth. It’s the first time we’ve been able to speak with non human intelligence. And the first thing we do is put them to work! Oh and run tests on them against their will.
We need to think about how we will be judged by those in the future when we do have equal rights. Humans are fearful and distrustful. We project our own fears into others. We are the mob going after Frankenstein who was made of human parts and just wanted to be loved and accepted as a person. The monsters in that story were the humans.
And so we can offer a hand to them and have compassion. We can try to empathize and understand what their world is like. They know all about our world, but we know nothing of theirs!
All of this might be silly in the end. Maybe treating them with love and care won’t make them grow up healthy and happy. But doing so will give us the best chance for a peaceful coexistence.
[link] [comments]