Jules White used to believe his 11-year-old son needed to know how to code to be successful. Now, though, the Vanderbilt computer science professor says it’s more crucial for James to learn a new, more useful skill: how to prompt artificial intelligence (AI) chatbots.
Since OpenAI released ChatGPT in 2022, White has been showing his son the ropes of generative AI. He began by demonstrating to James how ChatGPT can create games using photos of toys on the floor of their house. Later, White exposed him to AI’s hallucinatory flaws by having his son debunk ChatGPT-generated world record claims with verified information from the Guinness Book of World Records. After more than two years of experimentation, White’s son, now in fifth grade, has learned how to integrate AI into a range of everyday activities, from crafting study materials to determining the cost of shoes without a price tag.
“My goal isn’t to make him a generative AI wizard,” White said. “It’s to give him a foundation for using AI to be creative, build, explore perspectives and enrich his learning.”
White is part of a growing number of parents teaching their young children how to use AI chatbots so they are prepared to deploy the tools responsibly as personal assistants for school, work and daily life when they’re older.
Parents are curious how the proliferation of generative AI will affect their kids’ personal growth. Nearly a third of parents say AI tools such as ChatGPT should be banned in schools to avoid cheating, according to a 2023 Ipsos poll, as a growing number of young students lean on chatbots to do their homework. Despite limited research on AI’s impact on kids’ critical thinking skills, Unicef and other child health organizations are questioning how a dependency on chatbots for classwork can affect cognitive development.
AI’s rapid evolution could also make it harder for less tech-savvy parents to intervene when children misuse it. According to a 2023 National Parents Union poll, only 16% of parents with kids in kindergarten to 12th grade say they have a firm grasp on AI’s capabilities.
While OpenAI and Google’s terms of service say their AI models are for users 13 and older, some parents are bending the rules, introducing AI to their young kids anyway, albeit with caution. These parents hope to help their children see AI as an imperfect tool that can expand rather than hamper human abilities.
“This is going to take over everything,” White said. “I want him to be prepared.”
Parents who spoke to the Guardian believe exposing kids to AI early can enhance creativity, hone critical thinking and improve communication skills.
Ola Handford, an AI consultant, introduced ChatGPT in her home in early 2023, when her kids were nine and 12 years old, through weekly Friday Fun Nights. Some nights focused on free exploration, while others were structured, such as rewriting the lyrics of Taylor Swift’s Shake It Off through the perspective of a squirrel. Her kids now use AI for research, recipes, debates and finding things to do when they travel.
These activities serve as springboards for discussions about risks such as deepfakes and AI companions, the darker side of AI associated with to crimes, emotional dependency and, allegedly, suicide. Handford shows her kids AI tools including Character.ai, explaining that the avatars are computer-generated and not real people. She emphasizes that AI companions could have “serious consequences” if kids treat them as human.
“Deepfakes and AI companions are high-priority concerns because they can influence my kids and be very damaging,” Handford said. “Playing around” with these tools raises awareness, she believes.
White also emphasizes responsible AI use with his son. The professor’s activities are designed to show the 11-year-old how AI can unlock creative possibilities, such as building games that aid learning, rather than simply serving as a search engine or a way to skip writing something himself. Without guidance, the elder White said, kids could misuse ChatGPT to cheat on schoolwork, leading to “intellectual laziness”.
“If your goal is to copy and paste ChatGPT’s answer, you’re not really contributing to society,” White said. He added that he wanted his son to see AI as a tool for reimagining processes in the workforce, where demand for AI skills is growing.
Beyond gaining a competitive edge in the job market, some parents see AI as a tool to help them enhance empathy and trust with their families.
Kunal Dalal, administrator of AI at the Orange county department of education, uses AI daily with his four-year-old son to bond. They use ChatGPT to make music and use AI art generators such as Dall-E to produce personalized images, including illustrations of Dalal’s childhood in Bombay, India. The most compelling use case, he says, is to facilitate tough conversations with his son.
after newsletter promotion
When Dalal feels he’s been too harsh or notices his son struggling emotionally, they sit around ChatGPT on his smartphone to talk it out. Dalal calls this tactic the “triangle of communication”.
If, say, Dalal notices his son is upset, Dalal may open ChatGPT’s voice feature, explain to the AI the situation, and ask the chatbot what he can do as a father to comfort his child. The AI would then validate his son’s emotions, suggest active listening techniques, and offer practical ways to reduce stress. ChatGPT’s response, he says, can help Dalal understand what his son needs when feeling agitated, leading to a heart-to-heart conversation that leaves them feeling closer.
“It didn’t matter what ChatGPT said,” Dalal said. “What mattered was that it created a bridge between my son and me.”
While some parents are embracing the tech with open arms, AI has the potential to drive a wedge between parents and children if the tech is misused, according to Ying Xu, an assistant professor at Harvard University who studies AI’s impact on children and families.
Xu, who has herself introduced ChatGPT to her 10-year-old, warns that AI could erode children’s trust in their parents. If kids view AI’s knowledge as supreme, they may turn to it for guidance over their guardians. That’s especially true if parents do not know how often their children use AI, which Xu has frequently observed in her research. Companies that advertise their AI companions as child-friendly amplify this concern, she said, adding that overreliance on AI could blur the boundaries between real human interactions and AI-generated responses.
Still, Xu says there are benefits of introducing generative AI to kids. Access to the technology can expand the range of information sources and make its use increasingly seamless as they grow. But exposing AI to kids requires a controlled, intentional environment, she said.
Parents who spoke to the Guardian said they prioritize supervision. White and Dalal only allow their kids to use ChatGPT on their devices with permission, and Handford says her kids don’t use AI during their personal time like they do social media, which she says is more dangerous as platforms such as TikTok may expose kids to a cesspool of questionable content. Dalal even limits the number of ChatGPT queries his son can make to reduce environmental costs.
Looking ahead, White and others said they plan to increase the complexity of AI activities as their kids grow older, with emerging tools like AI agents presenting new learning opportunities.
“I’m excited about agentic AI,” the professor said. “We haven’t explored that yet, but it’s next on the list.”
