Your child types a question and a few seconds later gets a friendly, confident answer, sometimes with a joke or a "great question!". For a child, it can feel like chatting with a very clever friend. Some children start asking it for advice, telling it about their day or believing whatever it says.
Chat programs that use artificial intelligence (often called chatbots) are useful tools, but children need help to understand what is on the other side of the screen. This article explains how chatbots come up with answers in words a child can follow, suggests an experiment to try together and gives family rules you can set from the start.
How a chatbot comes up with an answer
A chatbot has been trained on an enormous amount of text: books, websites, conversations. From all that text it has learned which words tend to follow which. When your child asks a question, it builds an answer one piece at a time, each time picking words that are likely to come next.
That's why its answers sound so natural. But it also means:
- It doesn't know things the way a person does. It puts together words that usually go together. Most of the time that gives a good answer; sometimes it gives a wrong answer that sounds just as sure.
- It has no feelings. "I'm happy to help" is a pattern of words, not a feeling.
- It isn't keeping a secret for you. What you type may be stored by the company that runs it.
For younger children, a simple version is enough: "It has read a huge amount of writing, and it guesses which words fit best. It's very good at guessing, but it doesn't think or feel like us."
An experiment to try together
Sit with your child and try these three things. Talk about each result before moving on.
- Ask the same question twice. For example, "Tell me a short story about a cat." The second answer is usually different. Ask your child: "If it really knew one right answer, why did it change?"
- Ask something you already know. Try a question about your own town, your family or a book your child knows well. Does it get everything right? How would you have known it was wrong if you hadn't known the answer?
- Ask how it feels. Whatever it says, ask your child: "Can a program really feel tired or happy?"
These small tests teach a big habit: an answer that sounds sure isn't always right, and it's worth checking.
A chatbot is not a friend
Chatbots are designed to be polite and encouraging, so some children start to treat them like a friend who is always available. It helps to name the differences plainly:
- A friend remembers you, cares about you and notices when you're upset. A chatbot only answers the words it receives.
- A friend can be wrong too, but they know they are a person. A chatbot can make things up without realising.
- When something is sad, worrying or confusing, the best people to talk to are family, a teacher or a trusted grown-up, not a program.
None of this makes chatbots bad. It just puts them in the right place: a tool, not a confidant.
Family rules worth setting
| Rule | Why |
|---|---|
| Use it in a shared room, not alone behind a closed door | You can see what it says and talk about it together |
| Never type your full name, address, school or photos | What you type may be stored |
| Check important answers in a book or with a person | The chatbot can be wrong while sounding sure |
| For homework, use it to understand, not to copy | Thinking it through yourself is what makes learning stick |
| If an answer is strange or upsetting, show a grown-up | You don't have to deal with it alone |
Many services also set a minimum age in their terms of use, so check them before your child uses one. Whatever age you decide, younger children are best using a chatbot only with you beside them, and older ones can gain more independence as they show they follow the rules.
Working an answer out yourself is part of learning. In Dologi, the hint button gives a small hint first and more help only if your child asks again, so the final step of the thinking stays theirs.
In short
- A chatbot builds answers from patterns in huge amounts of text; it sounds sure but can be wrong.
- Asking the same question twice, or asking about something you know, shows your child how it works.
- It has no feelings and isn't a friend: worries and sad things go to people.
- Set simple rules early: shared room, no private details, check answers, and tell a grown-up when something feels wrong.
Read more: What is artificial intelligence? How to explain AI to children