Most talk about AI is about what it can do. The more important question is often what it should do, and children have sharp instincts about this. Separating “could” from “should” gives them a simple tool they can use for the rest of their lives: just because a machine is able to do something does not settle whether it is right to hand the job over.
“Will robots take everyone’s jobs?” Machines are taking over some tasks, and they will change many jobs. Nobody knows exactly how much. But it is not only up to the machines. People decide where machines are used, and those decisions can be argued about, voted on, and changed. That is part of why it is worth thinking about now.
]]>When people say a computer system should be “fair,” they often disagree about what fair means without noticing. Children already have strong feelings about fairness, especially when food is involved. Sharing the same snack three ways makes the disagreement visible, and lets everyone see that each rule helps someone and leaves someone else feeling short-changed.
“Can’t a computer just work out what’s fair?” A computer can follow a fairness rule very carefully, but it cannot choose the rule. That choice is about values, about what we think matters most, and people have to make it. When a system is called fair, it is always worth asking: fair in which way, and who decided?
]]>Chatbots and the word suggestions on a phone keyboard work by predicting what is likely to come next, based on enormous amounts of text they have learned from. Children are excellent predictors too. Playing with familiar and unfamiliar sentences shows them, from the inside, what prediction is good at and where it goes wrong: it is great at what usually happens, and can be confidently wrong about what actually happened.
“So does a chatbot actually know things?” It has learned patterns from a huge amount of writing, so it is often right about common things. But it is built to produce a likely-sounding answer, not to check whether the answer is true. That is why it can sound completely confident and still make things up. Checking with a book, a teacher, or another source is a good habit.
]]>Many systems that make decisions about people will not say how they decide. The only way to understand them is to try things and watch what happens. That is exactly what your child does here. They also run into a deep idea on their own: a guess can fit every example so far and still be wrong.
“Why wouldn’t a machine just tell you its rule?” Sometimes the company that made it keeps the rule secret. Sometimes the rule is so complicated, learned from millions of examples, that even the people who built it cannot say it in one sentence. People are still arguing about how much a machine should have to explain when it decides something about you.
]]>Automatic systems are easy to miss because they are designed to be. Before a child can ask whether a system is fair or useful, they need to notice that it is there. Sorting what they find into four kinds of job also teaches something subtle: a door that opens by itself is not making the same kind of choice as a machine that decides who gets a discount.
“Are all these machines AI?” No, and that is a good thing to notice. Many are simple: a sensor and a rule, like “open the door when something is close.” AI usually means a system that learned its rule from lots of examples instead of being told it. From the outside they can look the same, which is one reason it is hard to know when AI is being used.
]]>AI systems are built from data, and much of it comes from ordinary moments: a library card, a school sign-in, a loyalty card at the shop. A week of noticing turns an abstract word into a list in your own handwriting. The third column, “Did we choose?”, is the heart of the activity.
“Is it bad that people collect our information?” Not always. A doctor needs to know your health history, and a library needs to know who borrowed a book. The questions worth asking are whether it is needed, who else can see it, how long it is kept, and whether you were told. Those questions matter more as more decisions are made by computers using that information.
]]>Sorting is one of the main jobs AI systems do: this email is spam, that photo is a cat, this application goes to the next round. A sorting rule feels perfect until something arrives that it was never designed for. Children see this instantly with laundry, and the question they end up asking (“where does this one go?”) is the same question people ask about real systems.
“Do machines have a ’not sure’ pile?” Some do. A well-built system can say how confident it is and pass the hard cases to a person. Not every system is built that way. Some just put every item in a basket, even when the rule does not really fit.
]]>Adults fill gaps without noticing. Told “put jam on the bread,” we open the jar, pick up the knife, and spread. A machine does none of that unless it is told. When you play a robot that does only what it hears, your child discovers the gap for themselves, and they discover it by laughing at you holding a closed jar against a slice of bread.
“Are computers really that silly?” Computers are very fast and very exact, but they do not guess what you meant. People called programmers write the steps for them. Newer AI systems can follow looser instructions, because they have learned from lots of examples what people usually mean. They still get it wrong sometimes, in ways that can be surprising.
]]>AI tools can now make text, voices, and pictures that look real. The skill that protects children is not spotting tiny glitches, which keep getting harder to see. It is the habit of asking “how could I check?” This game builds that habit with stories about your own day, where everyone can actually find out the truth.
“Can AI make fake pictures?” Yes. Some AI tools can make pictures, voices, and videos of things that never happened, and they can look very real. That is why it helps to ask where something came from, whether anyone trustworthy is also reporting it, and whether it is trying to make you feel something strongly before you have had time to think.
]]>A shop is a feed you can walk through. Where a product sits is often planned, and some shops are paid by makers for the best spots. Feeds, app stores, and search results work in a similar way: the first thing you see was put there for a reason. A child who has crouched down to look at what is placed at their eye level has a real, physical sense of what “ranked first” means.
“How does a video app decide what to show me?” Usually a computer program ranks thousands of videos and shows the ones it predicts you will watch longest. It learns from what you, and people like you, watched before. That is not the same as showing you what is best for you, and the people who run the app choose what it aims for.
]]>