Should you ask AI for parenting advice? | Canada Voices

Alison Fleming has a confession: She recently used ChatGPT to potty train her four-year-old son.

“I have to say, it was a great resource for us,” says the Hamilton, Ont.-based marketing professional. “There were so many ways to make it seem fun—like pretending he was having a party—that really helped solve what had been a three-month struggle.”

She’s not the only parent turning to Claude, ChatGPT or another artificial intelligence (AI) source for advice about her kid.

In the age of the large language model (LLM), parents are asking for information about everything from birthday party ideas, health ailments and even the kind of “Am I even doing this thing right?” support you might seek from a therapist.

In fact, about 79 per cent of parents with children under the age of 18 use AI, finds a recent report from Menlo Ventures, a U.S.-based venture capital firm.

But, while AI can be a boon for a burned out parent seeking quick information and is ideal for generating creative, fun ideas, it can’t be completely trusted as a source of sound parenting information, says Keri Ewart, an assistant professor in the education program at Wilfrid Laurier University in Waterloo, Ont.

“A lot of people will use it to just get ideas and that is fantastic,” she says, “because it’s something that maybe you haven’t thought about or don’t know how to navigate.”

But she urges parents to use critical thinking above all else when searching parenting guidance. “To off source all of the ideas to ChatGPT– that’s not the way we want you to use it,” she says.

A confusing time

The duality of AI, its usefulness as a learning tool versus its untrustworthiness, has many parents wondering how best to use it. That’s understandable, says Dyan Eybergen, a registered nurse, behavioural health specialist and certified parenting coach based in Mindemoya, Ont.

“If we have parents that are relying solely on this, bounce it off your parent antenna,” she says. “Do what feels right to you.”

She says AI lacks the ability to consider individual child traits, which can be crucial for effective parenting.

“When we’re talking about asking it to give us strategies for a child, this is where I have a lot of concern because when we are parenting, it cannot be a one size-fits-all paradigm,” she says. “We have to take into account that child’s temperament, that child’s personality.”

At the same time, sometimes AI’s advice can be downright wrong.

She cites a recent example where a friend’s teenage daughter, an instructor at a fitness centre, had been approached by a parent seeking information about the times and locations of the lessons.

“This person was going through a very nasty divorce and the other parent had signed the kid up. The other parent had no information, so they wanted this teenager to tell them when the lessons were,” she says.

But when the teen asked ChatGPT for help in solving the situation, it returned an answer that was not appropriate: that she “should ask for the custody agreement,” she says, laughing at the ridiculousness of this incorrect advice.

Eybergen says parents need to trust their intuition and understand their child’s unique needs, rather than relying on AI-generated information that might be generic.

For Ewart, the most significant dangers of using AI to parent are when parents rely on AI-generated online videos for what they believe are educational purposes.

“A lot of parents are using YouTube to manage their children, rather than to teach their children,” she says. “They think, ‘Oh, this looks like a great video; we see the alphabet, we see numbers.’ But because this is AI there are so many hallucinations happening, and nobody’s vetting this whatsoever.’”

In many cases, young children are being exposed to misinformation while their speaking, reading, and thinking skills are slowly being eroded, she says.

Ewart suggests parents carefully vet any educational videos prior to letting their children watch them and to always look at who has produced the video, ensuring that it’s a reliable source, such as TVO, CBC Kids or PBS.

A critical eye and common sense are truly the best ways to approach AI parenting, she says.

“If parents are choosing to have their children watch these videos and then themselves engage in AI as well for advice as parents, do so in a really safe and responsible way.”

Better yet, she recommends parents forgo videos and spend as much time possible interacting with their children one on one, whether it’s reading or playing in the backyard.

It’s for these types of activities that Claude or ChatGPT might actually have some great ideas.