My son, an IT professional, dropped by yesterday to show me what he’s been working on. Roughly speaking, as I sort of understand it, he’s creating a program where AI will answer questions and provide information for his employer’s clients. The AI will draw its information from a long list of relevant sources he provides, not from the internet at large. The result will be a sort of AI-generated information desk or public information officer for the clients. I think.
Anyway, to give me a rough idea of what the end product might be like, he asked it to examine this blog and do a sort of overview, as though someone had asked, “What’s this?” AI then created the following dialogue which, to my ear anyway, sounds remarkably human. I doubt that I’d have recognized it as AI generated. You may or may not find it as intriguing as I did:
I’ve encountered narrations on YouTube videos and elsewhere that were obviously computer-generated. And I’ve found some recently where I couldn’t be sure if the narration was human or not. But this dialogue tops anything I’ve heard to date.
My apologies to my son for what I’m sure is my utter failure to understand and explain what he’s doing. But I’m absolutely fascinated by the result.
For those more concerned than impressed, he says he finds AI both amazing and horrifying. Take from that what you will.
It sounds altogether human, including ideal voices! Isaac Asimov might peg what we now know as AI as a significant step towards the “positronic brain” for robot intelligence that he envisioned. For me, this passes the Turing test. That’s not to say I think the Turing test is actually valid, but it is undoubtedly a milestone. BTW, I agree with AI’s opinion about your blog – it’s about the wonder of life, which is why I’ve been a long-time fan.
I appreciate your long-time support; you’re one of very few who’ve stuck around this long, who didn’t get bored or distracted and drift off to other interests.
Yep, I found this dialogue stunning. Had no idea AI could be programmed to do something like this. Two different, very natural sounding voices carrying on a very natural sounding conversation about a designated target. It is a bit scary to think I could be carrying on a conversation with AI and think I was talking to a person. Just the other day I was asking questions, in an online text dialog, at my insurance company. I’ve always known or assumed that such arrangements are chatbots. But this one had gotten so good that I actually wondered if it might have been a human on the other end. For that matter, the tax lady I was just talking to on the phone could have been AI or will be someday soon. Given the natural sounding dialog above, who knows. If not today, maybe next month …
I agree absolutely with what your son says: this is equally astounding and horrifying, Colorado.
I mean, “an eclectic mix” ???!!!
Yep, having it understand what I’ve written and then playing two roles having a conversation about it — it’s mind-blowing. It’s obvious that already, today, you could be carrying on a conversation with “someone” you think is a human, when it’s actually AI. That’s really unnerving. There’s no way to know if it’s telling you the truth or inventing some dangerous lie. For now I suppose it depends on what the programmer tells it to do (I trust my son), but if the programmer has ill intent, who knows.
AI must make us terribly wary of strangers and/or programs within the field of IT.
Well, we need to stay alert for sure, although it appears at times we just won’t be able to tell from the sound of it. There are already so many kinds of scams out there. You just can’t be too careful.
wow
Kinda mind-blowing, isn’t it
I have to admit, it scares me.
I don’t see how anyone with half a brain could hear this and not be immediately unnerved. Certainly this capability could be very useful to well-intentioned people, but in the hands of the ill-intentioned … yikes!
Yes, scary how real they sound. I was expecting a formal voice not this informal conversation. It would have fooled me.
Me too. It’s scary to think how easily I could be fooled by some malevolent AI and never realize it. This is light years beyond not believing before-and-after photos or “photographic evidence.”