ChatGPT ‘constantly asked’ Sam Altman about his newborn
Over the course of our existence, we have been unable to answer a simple question: Why is my baby crying?
Sam Altman is the father of a three-month-old baby and CEO of OpenAI. He hopped on OpenAi’s podcast to discuss how his company has impacted his experience of fatherhood. Altman, who calls himself “extremely child-pilled,” says he “constantly used ChatGPT” to ask questions about baby behavior during the first few months of his son’s birth. Now that he is more settled, Altman uses ChatGPT for more general questions regarding children’s development stages. Altman stated that people have been taking care of babies for a very long time without ChatGPT. “I don’t know how I would have done that.”
It’s not fundamentally different than frantically Googling baby questions, something even the best-prepared parents do for decades. Altman’s choice of internet tool is not surprising, but given his background, it is.
However, since hallucination is still a problem for AI products, it could be unsettling to rely so heavily on a chat AI when it comes to baby care.
Parents have been known to seek out information from questionable sources in the middle night. My colleagues who have children describe Google as a “bottomless pit”and parenting Facebook groups as a minefield. Is ChatGPT any different from taking advice online from someone who insists that you’re a negligent caretaker if your baby’s sleep time isn’t based on the current moon phase? Altman discussed the idea that parents using AI to find child-raising solutions is less alarming than the idea that very young children use it.
Altman said, “There’s a video that has always stuck with me. It shows a baby or toddler with an old glossy magazine [tapping] and the [cover].” The child believed that the magazine was an Apple iPad. “Kids will think that the world has always had very smart AI.”
Former OpenAI scientist communicator Andrew Mayne was interviewing Altman and recalled a social media posting from a parent using the voice mode of the ChatGPT app to talk to their child about his obsessions.
Mayne giddily said, “He got tired talking to his child about Thomas, so he put ChatGPT in voice mode… An hr later, the kid is still talking about Thomas, the train.”
“Kids love voice mode,” Altman interjected. As parents use ChatGPT to do similar things, the discourse will likely reflect the same “iPad kid generation” (yes, you shouldn’t let your child watch “Cocomelon”but it’s also not fair to expect that parents occupy their children’s time 24/7).
Existing children’s media, however, is created by humans. ChatGPT, on the other hand, has its own policies. We do not recommend that children under 13 years old use it. It doesn’t have a parental control mode that has been vetted. He said that Altman was aware of the risks.
It’s not going to be all good. Altman said that there would be problems. Altman is right. “People will develop parasocial relationships that are problematic or even very problematic. Society will have to find new guardrails.” We don’t know the full impact of letting children talk to a language model for an hour about Thomas the Tank Engine. Altman is the CEO of a huge company that spends billions on AI. He never forgets this in his messages. Altman said, “The upsides are tremendous!”
“Society is good at finding ways to mitigate the downsides.”
Amanda Silberling, a senior writer for TechCrunch, covers the intersection of culture and technology. She has written for publications such as Polygon, MTV and the Kenyon Review. She is co-hosting Wow If True, an internet culture podcast, with science fiction writer Isabel J. Kim. She worked as a museum educator, film festival coordinator, and grassroots organizer before joining TechCrunch. She holds a B.A. She holds a B.A.
Send tip through Signal, a encrypted messaging app, at (929) 593 0227. For anything else, email [email protected].
View Bio
