Sam Altman's ChatGPT parenting hacks have stirred up a hornet's nest. - 36氪

Sam Altman's ChatGPT parenting hacks have stirred up a hornet's nest. 量子位 2026-08-10 07:52 Was utterly blown away by a single remark from the core creative team of Disney Animation. Sam Altman has been roasted relentlessly again, this time he was excitedly trying to teach everyone to raise kids with GPT... Recently, Sam Altman, co-founder and CEO of OpenAI, shared a ChatGPT use case he believes is really cool on X: Input all the family's schedules and children's interests into ChatGPT, then let it automatically generate an exclusive family podcast on the way to send kids to school every morning, chat with the second child about the afternoon football match, talk with the eldest child about the upcoming birthday, and smoothly intersperse a few news pieces for the adults in between. Sounds like every big and small matter of the whole family is arranged perfectly by ChatGPT, and parents no longer need to rack their brains to find common topics with their children. However, as soon as the post was published, the comment section exploded. Some netizens raged "How dare you think like that!" ; some netizens were in disbelief "This is the most anti-human AI usage I have ever seen!" ; others taunted: "Why don't you let GPT read it out loud to you before you post? Isn't that even cooler (doge)." Alex Hirsch, the creator of the Disney animated series *Gravity Falls*, even delivered a fatal blow with just one sentence: Actually, you can just talk to your kids directly. As of press time, Sam Altman's original post has received about 11,000 likes and 3,900 reposts ; Hirsch's retort has gained about 202,000 likes and 17,000 reposts , with the number of likes 18 times that of Sam Altman's post. This is rather awkward. This CEO, who once stated painfully in a Relentless interview that "not having time to spend with my children is the most painful thing for me", now proposes the solution of "letting AI chat with the kids on my behalf", the contrast between his public images before and after is extremely fragmented. What netizens really mind is not that Sam Altman uses AI to organize family schedules, but that AI starts to decide for parents what to talk about and how to talk on the way to school, and even directly processes parent-child conversations into a formal program. This is very wrong. Sam Altman's AI parenting idea backfires In fact, this is not the first time Sam Altman has seriously promoted ChatGPT for parenting. In June 2025, he revealed on the official OpenAI podcast that he kept asking ChatGPT all kinds of questions about his son, including how children should develop at different stages. Sam Altman once said frankly: I know parents have raised countless generations of kids without AI, but as a new dad, I find myself dumping all those 2 a.m. questions that I would otherwise bother the pediatrician with to GPT. In December of the same year, Sam Altman appeared on *The Tonight Show* hosted by Jimmy Fallon, and talked about ChatGPT parenting again: I can hardly imagine how I would raise a newborn without ChatGPT. But he quickly added another sentence: Of course, it's totally fine not to rely on AI. It's just that I rely on it a lot. He gave an example: his child could not crawl at the age of 6 months, and he was extremely anxious for a while, so he turned to ChatGPT for help. The AI told him that every child has a different development pace, and this situation is not necessarily abnormal, so he finally felt relieved. At this point, ChatGPT is more like a on-call parenting encyclopedia . For flustered new parents, when encountering problems like children crying, having a fever or developmental issues in the middle of the night, being able to quickly get some basic information is indeed much more convenient than flipping through dozens of web pages in the search engine. The image Sam Altman shaped from this is also a very specific new dad: he gets anxious, gets overwhelmed, and will turn to his own company's AI products for answers. This gentle dad filter used to be quite stable, and it was even often talked about enthusiastically by the media. Until this time, he put the family schedule, children's interests and chat content on the way to school all into ChatGPT, the filter began to shatter. Everyone finally sees clearly that he is first and foremost the operator of OpenAI, and a father second. Making full use of AI to remind us not to forget that there is a match in the afternoon is certainly great. But if even asking the child whether he had fun playing football and how many goals he scored has to be done by AI on your behalf, something starts to feel off. This is not just a dad's random idea, OpenAI is really targeting the whole family What's more noteworthy is that this is not just a temporary random thought of Sam Altman as a new dad. They have already started recruiting people. According to TechCrunch reports, OpenAI posted a job opening in July to recruit a product manager dedicated to family products . (The position has been removed now, presumably they have already hired someone.) It is not hard to see that the positioning of ChatGPT is quietly shifting: in the past, it mainly served personal learning and office work; but now, it is trying to become the "digital butler" of the whole family , coordinating the information and daily arrangements of multiple members. At the same time, the user structure of ChatGPT is also changing. Estimates provided by market research firm Sensor Tower to TechCrunch show that in the second quarter of 2026, the proportion of ChatGPT global users aged 35 and above rose to 31% , a significant increase from 26% a year ago; the proportion of users aged 18 to 24 dropped from 34% to 29%. Among parent groups who use smartphones in the United States, the coverage rate of ChatGPT is about 24% , higher than 16% in the same period last year. This shows that ChatGPT is obviously no longer just a tool for students and young office workers, more and more parents are bringing it into their homes. And the "family calendar + children's interests + exclusive podcast" proposed by Sam Altman this time just shows the embryonic form of this kind of "family AI" : It not only knows what you are going to do tomorrow, but also knows what your children like, where the whole family will go next, and what things are worth reminding you of. This post of Sam Altman that was widely mocked may have just leaked OpenAI's ambition for the next stage: let ChatGPT not only know you, but also know your whole family. The cost of AI entering the family is more sensitive than that of entering the office But the problem is, family AI is completely different from office AI. If office AI makes a mistake, what gets ruined may only be a reproducible spreadsheet; but if family AI makes a wrong judgment, what gets affected may be the next generation's education, mental health and trust in their parents. Related risks are no longer at the hypothetical level. According to AP reports, OpenAI is currently facing multiple related lawsuits . In November 2025, seven families filed lawsuits against OpenAI, four of which alleged that ChatGPT played a role in the suicide of their loved ones, and the other three alleged that ChatGPT reinforced harmful delusions. Perhaps for this reason, OpenAI is continuing to advance family products while urgently adding safety guardrails. But OpenAI itself also acknowledges that technical guardrails are not foolproof , and the protection mechanism may still be deliberately bypassed, requiring continuous adjustment and improvement. Moreover, data privacy is also an issue that OpenAI cannot avoid. To realize the family usage scenario described by Sam Altman, we at least need to provide AI with all kinds of information such as family schedules, children's interests and activity arrangements. At the same time, as the functions continue to deepen, names, ages, school affairs, family relationships and shared memories may all become the context required for personalized services. Then how long should this information be kept? Do all family members agree to upload it? Can the uploaded family information be guaranteed to be safe and not stolen? When the child grows up, can they delete the records left by their parents on their behalf? It's hard to guarantee, right? The smarter the family AI is, the more information it needs to know; the more it masters, the heavier the responsibilities for security, privacy and ethics will be. So at least at this stage, no matter how perfect the technical guardrails are, they cannot replace the sincere communication between parents and children. Clear boundaries are needed when AI enters the family Of course, Sam Altman's suggestion is not completely useless. For parents who are occupied by picking up and dropping off kids, busy work and all kinds of housework, AI that can automatically organize the whole family's schedule and remind them of important matters can indeed help them reduce a lot of burdens. But the rhetorical question from Alex Hirsch still points out the most sensitive boundary of family AI: Some things can be handed over to AI, while some things need parents to complete in person. AI can remind you that your child has a football match in the afternoon, sort out the rules of the game, and even prepare a few chat topics; But it cannot ask "Did you have fun playing today?" on your behalf, let alone listen carefully to the child's answer for you. As this netizen said, AI cannot replace a mother to shed tears at 2 a.m. when her child has a fever ; cannot 该文观点仅代表作者本人，36氪平台仅提供信息存储空间服务。