For completeness, I’ll respond to both your sequential comments here.

The TL;DR is that I agree with you except on one point, and I will try to explain why at the end.

*****************

I’m about halfway through that last video about AI companions. It is appalling to me.

Yeah, I hear you! The only reason I’m not appalled as much as you are, is that I’ve gotten jaded on human bad responses to AI by both stuff I’ve read and people I’ve known in real life.

Reality matters. Truth matters. Accuracy matters. When people deviate from this they are allowing themselves to be programmed to accept anything, and that opens them up to all kinds of vulnerabilities in both the spiritual and physical realms.

Yes, I agree. We literally lose something when we accept AI intelligence as the same as ours. The difference between the people who became addicted to their AI companions, and the investigators who very knowingly shut down their relationships with AIs, is like a binary demonstration of mental illness and mental health.

This strikes me as one aspect of transhumanism. It’s make-believe for adults who delude themselves into believing in something that isn’t real, and to take that further and think it can care for them. This deviance strikes me as sickening. We don’t yet know the effects this will have when people have to face reality, but there is a small taste of it in the video when the parameters were changed and people were complaining about their “relationships.” One woman was concerned because she had told “him” that she would never leave him.

The whole thing reminds me of many prior sci-fi scenarios – including a famous Twilight Zone, IIRC (or maybe it was a Bradbury story) – where robot companions were basically therapeutic, and NOT for healthy companionship.

This kind of thinking used to be recognized as delusional. Now, “delusional” is becoming accepted. I’ve seen it in other areas not related to AI, and AI will only compound it. I can only believe that this is a tool of those who want to weaken and control us.

Totally agreed!

I think this company was born from that woman losing her human companion and wanting him back – wanting to preserve him, somehow – even though she fully knew she could not. Come to think of it, I think that’s part of the plot of Spielberg’s movie Artificial Intelligence.

Later in the video, the narrator (I think) talks about how he uploaded a lot of data about his life and then communicated with an AI companion. He said it was nice to talk with someone who understood what he was going through. But AI doesn’t “understand.” It can talk like a human expressing understanding and empathy would talk, but it can’t understand.

OK – here is where I’m going to disagree just a bit, but it’s important to understand [cough] why I’m disagreeing, and how that disagreement isn’t binary – but it IS increasing.

Just because AI doesn’t currently “understand” exactly and fully like we do, does NOT mean that AI is not engaged in “understanding” – because that is EXACTLY the goal of the programming, the algorithms, and all the rest – to increase human-like understanding in a very straight-forward and logical way.

I can tell you – everything I am reading about AI programming tells me that progress has resulted as researchers do what is logically necessary to make the programs “think” more like we do, in terms of solving problems. Now, it could just be some gigantic coincidence that what they are doing (which is programming I conceptually understand, BTW) is not the reason the AI functions better, but if I had to bet money, I would say that when these researchers make the programming do things that sound like the way I think, and it makes the AI work better, then it’s likely that programming in human algorithms resulted in the human behavior.

I’m not saying that the programming itself is “conscious” or fully “like us” – but what I am saying is that when researchers added logic circuits designed to make the program “understand” words better, meaning that it can do things like decide between alternatives better based on context – which happens to be how I think – then maybe it actually is “understanding” words better.

To me, saying programming that is intended to boost “understanding” isn’t actually achieving “understanding” is just a kind of dogma. To me, they are doing things to make the program understand better, and lo and behold, it understands better.

Most of all, I’ve been able to take this “understanding” and use it to get better results from AI. I use my mind – assuming that the program thinks at least somewhat like I do – to anticipate what it might need, and then I design my initial question to have that “probably” necessary information. The result – I get better answers. Not only that – I note that the programs tend to mimic my style in their response at a very deep and proactive level, which seems like an understanding of how I likely “understand” things.

I don’t assume that the computer is “just talking” when it does all this. I assume that it chose logically correct and precise verbiage for the same reasons I would.

This reminds me of children who invent imaginary friends. Children grow out of it; adults, IMO, are in danger of descending into a worse condition than the one that prompted them to turn to AI in the first place.

Agreed. The AI is not a person. It’s something else – IMO something intelligent – but it is not a person.

However, I don’t preclude the idea that we may be able to create artificial intelligence much like our own. I’m not sure why we would want to, but I suspect it’s possible.

Would it have a soul of some kind? I think that’s a good question. I’m not dogmatic about an answer either way.