Sam Altman has appeared in the first episode of OpenAI’s brand new podcast, called simply the OpenAI Podcast, which is available to watch now on Spotify, Apple Podcasts, and YouTube.
The podcast is hosted by Andrew Mayne and in the first episode, OpenAI CEO Sam Altman joins the host to talk about the future of AI: from GPT-5 and AGI to Project Stargate, new research workflows, and AI-powered parenting.
While Altman’s thoughts on AGI are always worth paying attention to, it was his advice on AI-powered parenting that caught my ear this time.
You have to wonder if Altman’s PR advisors have taken the day off, because after being asked the softball question, “You’ve recently become a new parent, how is ChatGPT helping you with that?”, Altman somehow draws us into a nightmare scenario of a generation of AI-reared kids who have lost the ability to communicate with regular humans in favor of their parasocial relationships with ChatGPT.
“My kids will never be smarter than AI.”, says Altman in a matter-of-fact way. “But also they will grow up vastly more capable than we were when we grew up. They will be able to do things that we cannot imagine and they’ll be really good at using AI. And obviously, I think about that a lot, but I think much more about what they will have that we didn’t…. I don’t think my kids will ever be bothered by the fact that they’re not smarter than AI. “
That all sounds great, but then later in the conversation he says: “Again, I suspect this is not all going to be good, there will be problems and people will develop these problematic, or somewhat problematic, parasocial relationships.“
In case you’re wondering what “parasocial relationships” are, they develop when we start to consider media personalities or famous people as friends, despite having no real interactions with them; the way we all think we know George Clooney because he’s that friendly doctor from ER, or from his movies or the Nespresso advert, when, in fact, we have never met him, and most likely never will.
Mitigating the downsides
Altman is characterizing a child’s interactions with ChatGPT in the same way, but interestingly he doesn’t offer any solutions for a generation weaned on ChatGPT Advanced Voice mode rather than human interaction. Instead he sees it as a problem for society to figure out.
“The upsides will be tremendous and society in general is good at figuring out how to mitigate the downsides”, Altman assures the viewer.
Now I’ll admit to being of a more cynical bent, but this does seem awfully like he’s washing his hands of a problem that OpenAI is creating. Any potential problems that a generation of kids brought up interacting with ChatGPT are going to experience are, apparently, not OpenAI’s concern.
In fact, earlier when the podcast host brought up the example story of a parent using ChatGPT’s Advanced Voice Mode to talk to their child about Thomas the Tank Engine, instead of doing it themselves, because they are bored of talking about it endlessly, Altman simply nods and says ,“Kids love Voice Mode in ChatGPT”.
Indeed they do Sam, but is it wise to let your child loose on ChatGPT’s Advanced Voice Mode without supervision? As a parent myself (although of much older children now) I’m uncomfortable with hearing of young kids being given what sounds like unsupervised access to ChatGPT.
AI comes with all sorts of warnings for a reason. It can make mistakes, it can give bad advice, and it can hallucinate things that aren’t true. Not to mention that “ChatGPT is not meant for children under 13” according to OpenAI’s own guidelines, and I can’t imagine there are many kids older than 13 who are interested in talking about Thomas the Tank Engine!
I have no problem using ChatGPT with my kids, but when ChatGPT was available they were both older than 13. If I was using it with younger children I’d always make sure that they weren’t using it on their own.
I’m not suggesting that Altman is in any way a bad parent, and I appreciate his enthusiasm for AI, but I think he should leave the parenting advice to the experts for now.