Although it may seem slightly off-topic for my “auto-momming” blog, as a mom and a data scientist, I am deeply concerned about the impact of new technologies on our society, particularly our children.

ChatGPT is really buzzing these days, it’s amazing and capable of performing a wide range of tasks. A teacher friend of mine even uses it to create curriculums from scratch! We use it at work to prepopulate fields and automate repetitive tasks, and many companies employ it as a chatbot for customer service. However, we are human, and humans are always great at using technology for their end games, and that end game can be good AND bad.

I won’t delve into the current dangers of ChatGPT for kids because you can find a ton of articles on that already. Instead, I want to discuss future and potential hazards that our children must prepare for.

Gaming

You know how in open-ended games, like the ones where you talk to characters and ask questions to achieve your goals, you used to interact with human or bot characters. You would typically get a pre-selection of questions to choose from.

Now, thanks to ChatGPT, those questions are becoming just regular, open-ended conversations. What does this mean for our kids? Growing up with this technology, they could potentially miss social cues in digital environments and struggle to differentiate between people and bots. This creates a whole new set of problems, which leads me to my second point.

Online grooming

We have all heard about kids being groomed online for sexual exploitation or worse. There are incredible organizations dedicated to digital safety and the prevention of this (My hat’s off to you, White Hatters!). It is known that grooming is something that takes months, so it is a real time investment for predators. Now, imagine if they could automate this with AI like ChatGPT? They could create multiple threads of conversations with teenagers and kids to establish the trust and grooming needed to meet in person. We already know how ChatBots can be moody and manipulative without a predator behind it, imagine the possibilities when it’s trained to act a certain way…

Scams

This goes without saying, and I think (I hope) this is more for the elderly than for kids, but ChatGPT can kick scammers’ operations up a notch. Imagine you receive a Facebook message or text from someone who claims to be a friend of your mother, daughter, husband, etc. They are in trouble and need money. Throughout the conversation, they can refer to things that only a friend would likely know, like your favorite restaurant, flower, or middle name. It sounds legitimate, right? Well, that kind of information is readily available to anyone with good internet search skills these days, and AI is easily trainable.

So, what are the things we can do to prepare our kids for ChatGPT?

• Give them a third-party resource: The White Hatter is great for that in Canada, there is Cyberwise in US as well as iKeepSafe
• Step by step digital intro: Instead of giving them a phone at 12, get them ready slowly. 8 is a good age to start texting, you can limit it with parents only, then add one trustworthy friend so on so forth…
• Discover ChatGPT together: Play with ChatGPT together, point out how its answers are different than a human, try to type things together to make it reveal itself as a bot and how it behaves.
• Encourage the flipside: Support technologies that are working to recognize an AI text vs. human.

Other ideas? Let me know in comments!