I talked tech with third graders for 90 minutes. Here's what happened.

How career day with 8-year-olds turned into 90 minutes on algorithms, business models, and data. Recreated from memory with creative liberties.

“Raise your hand if YouTube has ever shown you a video you didn’t search for.”

Every hand in the room went up.

“Now keep your hand up if you ended up watching it.”

Not a single hand went down. Whispers and giggles scattered across the room.

We were 45 minutes into what was supposed to be a talk about my career, and I had completely abandoned my outline. Instead, I was standing in front of a room full of third graders talking about algorithms, why they see what they see on their screens, and how YouTube decides what to show them next.

And they totally got it.

Let me back up: Last week I spoke to my kid’s third grade class for career day. I was supposed to talk about my career path, but what I really wanted to do was talk with all of them about technology. Not about the future, but about how what is on their devices right now actually works, and what their own experiences are.

To my delight we spent the next hour on three main topics: (1) algorithmic mediation, (2) business models & incentives, and (3) data. Of course, highly over-simplified for an 8-year old audience, but (to my great surprise) the basic points landed.

Below is my somewhat dramatized synopsis of how it went.

1. Algorithmic Mediation (or… who is making decisions here?)

I started with a question: Do you like it when your parents tell you what to read, or who your friends should be, or what clothes you should like?

• Class: No!

• Me: Ok. Me neither! Does anyone know what an algorithm is?

A few hands went up. The basic answer that surfaced was some version of “code.”

• Me: Yes, in a way. Algorithms are basically rules for computers. Where do they come from?

• Class: The computer?

• Me: They are created by people, and they are created to accomplish something specific — they have a goal. Who here knows what YouTube is?

Every single hand went up.

• Me: What is YouTube?

• Class: Funny videos! YouTubers! Lego videos! Dinosaurs!

• Me: Who makes those videos?

• Class: … people!

• Me: Yes. All kinds of people, anywhere in the world can put anything they want on YouTube. How does what you see on YouTube show up on your screen?

• Class: [silence]

• Me: An algorithm decides. A group of people decided how the algorithm should decide what you see next, and those are the rules YouTube follows. So… did YOU decide what you see?

One little boy pipes up: But I find the videos myself…!

Ok, now we are cooking.

• Me: Sort of… how did you find it?

• Kid: I searched.

• Me: And then what happened?

• Kid: I saw more stuff like that.

• Me: How did YouTube know you wanted more of that?

• Kid: Oh! The algorithm!

Yes, small person. The algorithm.

I go on: And what does the algorithm want you to do?

Silence for a bit, and then a hand goes up: To keep looking at videos?

• Me: EXACTLY! To keep looking at videos. And how does it do that?

• Kid: … it… gives me more videos?

• Me: Yep. It gives you more of what it thinks that you want. And it can do this because it watches what you do — on YouTube and other places on the internet — and then based on its rules, it guesses what you want to see. It guesses what will keep you looking at videos. This is also what happens on TikTok and anything else powered by an algorithm.

A room of about 30 kids was totally silent. I asked: How many of you use a tablet or phone?

All the hands went up. Do any of you ever have a hard time putting it down and walking away?

Heads nodded. I ask: Now why is that?

A hand went up. The algorithm doesn’t want us to walk away?

I honestly almost cried. Correct. The algorithm doesn’t want you to walk away.

At this point I was ready to pivot to business incentives, but before I did, I recapped the discussion with what accidentally turned into a somewhat comical call-response:

• What is an algorithm? Rules for computers!

• Where do they come from? They are made by people!

• What are they created to do? To keep us using it!

I continued…. So remember: you said that you don’t like it when your parents tell you what to do, what to read, what to wear, who you should be friends with… but that’s exactly what algorithms are doing—it’s just invisible. What you see is mostly decided by the algorithm, but did you know that we are not even allowed to know how the algorithm works?

The ENTIRE class erupts. WHAT?! That’s not fair! Why is that allowed!?

They got it immediately. I responded: I know… Why?... Well… they do it because people who could stop them, let them. And then I moved on because honestly, I had no idea how to have that conversation.

Onto the business model…

2. Business Models + Incentives (or… why is addiction the goal?)

I called back to an earlier moment: Ok, so they want you to keep watching videos. Why?

No hands. I said: Basically the more time you spend on their app or site, the more money they make. And the goal of any company is to make as much money as possible.

I asked them about their own use. What do they spend time on their devices doing? They were VERY excited to share their answers: YouTube! Minecraft! Funny videos! Roblox! Amazon! AI picture apps!

We talked about why those things are designed to keep you coming back. Why games have daily rewards. Why videos autoplay. Why you get notifications. Why games ask to buy things to use in the game.

Have any of you played with a claw machine? Heads nod. Do you ever win? Lots of chatter, and the basic conclusion is not really but sometimes.

I tell them: Right. That game is designed to make you THINK you can win… “just one more try mom!! PLEASE!!!” (they laugh) so you stay there and keep putting money in and keep asking mom or dad for more money that the claw machine gets to keep, while you get frustrated and mad (more laughs). Games and apps powered by an algorithm are just like this.

Heads were nodding. Ok, I think they got it.

Now, on to one of my favorite topics… and I really didn’t know how this one would land, but somehow it did…

3. Data (or… who is actually in charge of your stuff?)

This one I knew would be tricky. So I started very concrete: Do you take pictures on your devices?

Every hand went up.

• Me: Who is in charge of those pictures?

• Class: Me! My mom! My dad! My sister!

• Me: Really? Where do those pictures live?

• Class: On my ipad, so they’re mine.

I paused. How many of you have heard of the cloud?

About half the class raised their hands. (Wait seriously?!) Ok, who can tell me what the cloud is?

The answers came in. Surprisingly, they were not wrong… so I validated, and then I said: the cloud is really just somebody else’s computer. So if you have an ipad full of pictures, where do your pictures live?

Silence… then a hand: on Apple’s computer?

YES. So who is in charge of your photos?

I could see them processing this… then another hand: … Apple?

YES. They are still your photos, but are you really in charge of them?

I tried an analogy they’d understand: If you let your friend borrow some of your toys and they take them home, they are still technically yours, but who is *in charge* of your toys?

• Class: My friend.

• Me: Correct. Your friend is now in charge and they can play with them as much as they want. Can they give your toys to other people?

• Class: ...Yes.

• Me: What if you tell them they cannot give them to other people? What if they promise they won’t let other people play with your toys?

• Class: [thinking…]

Right. So... who is in charge of your photos?

Hands shot up. The lightbulbs were going off. THAT’S NOT FAIR!! THEY ARE MINE! WHY CAN THEY DO THAT?

Excellent question. This didn’t compute with their sense of fairness, and they really wanted to know WHY they were allowed to do this. One kid suggested we just “don’t use Apple.”

• Me: That’s an interesting idea… what would you use?

• Kid: I don’t know!

• Me: what do you use to take photos?

• Class: My mom’s phone, my tablet…

• Me: Right. So let’s make a list of different tablets and phones.

I made a list on the board as they call out device names. Next to each I put the name of the company. Google, Apple, Samsung, Amazon… This conversation was FASCINATING. They were floored that there were only a couple of options.

I hadn’t planned to talk about monopolies, but that’s what ended up happening — and they totally got it.

But then just for funsies, because I was already floored by how much they could engage… I really really wanted to know what they knew about generative AI.

4. Generative AI

I asked: who knows what a chatbot is? Nothing. Ok. Who knows what ChatGPT is?

Almost every hand went up. A tool! A robot! (lots of laughs for some reason)

• Me: ChatGPT is a type of chatbot… A chatbot is a thing you type into and it responds back to you. It’s not quite a tool — a tool is something that humans can control and use in a specific way. This is a bit different. Is it like a calculator?

• Class: Yes!!

• Me: Is a calculator ever wrong?

• Class: …no.

• Me: If everyone in this class types 20 + 20 into a calculator, what will the answer be?

• Class: 40!

• Me: Right. Every single one of you would get 40. But that’s not how a chatbot works. If everyone in this class typed exactly the same question into a ChatGPT, you would each get different answers.

The class erupted. I continued: Who knows what autocomplete is? A few hands went up.

• Kid: it helps me type a text on my mom’s phone.

• Me: Yes… it corrects misspelled words or fills in words… Is it always right?

• Kid: … sometimes.

• Me: Right. Do you know how it works?

I went on to explain that autocomplete is basically trying to guess what they want to say based on patterns it has “learned.” And then I said: That’s kind of how chatbots work. They are trying to guess what you want. How do chatbots guess? Silence

• Me: I’ll give you a hint: it’s similar to how YouTube decides what videos to put on your screen.

• Class: AN ALGORITHM!!!!!

• Me: Yes, exactly. So… what does ChatGPT want?

• Class: For you to use it all the time.

• Me: Correct. And how does it do that?

• Class: By giving you stuff you want.

• Me: Correct. Does the stuff you want need to be correct?

From here we talked about how chatbots are programmed to give you what you WANT, and not necessarily what’s correct or good for you.

• Me: Who likes candy? What’s your favorite?

• Class: Reeses! Skittles! Nerds! Twix!

• Me: Should you eat them every day?

• Class: laughs.

• Me: Who stops you from eating candy all day?

• Class (grumpily): Mom… dad… my dentist haha!

• Me: Why don’t they let you eat candy all the time?

• Class: … … … because… it’s not good for us. It ruins our teeth. It’s too much sugar.

Exactly.

• Me: Right. It’s not good for you, and the people who care about you will sometimes have to stop you from doing what you WANT because sometimes what we want is not good for us. What does a chatbot want?

• Class: … for us to use it a lot.

• Me: Why?

• Class: to make money.

• Me: What makes a chatbot work?

• Class: … an algorithm?

• Me: EXACTLY. And what is an algorithm?

• Class: rules for computers!

• Me: Where do they come from?

• Class: People!

• Me: What is their goal?

• Class: For us to keep using it!

A+ small people. I closed by saying something to the effect of:

Many of these things we use on screens can be useful and interesting — but WE need to be in control of them, and WE need to be making decisions for ourselves… So when you use your ipad, or play Minecraft or whatever… remember that the technology you use doesn’t care about you. The company that makes that technology doesn’t care about you either. They have made a thing that is useful in some ways, and their job is to make money… YOUR job is to make your brain as strong as possible, so you can decide for yourself how and when and for what you use all of this stuff… and for now that might mean reading books and doing your homework and listening to your parents when they say it’s time to turn off Minecraft. (laughs)

3rd graders are totally capable of engaging on these issues.

I was surprised that I didn’t have to convince them that something is off. They already feel it. They just don’t have the language for it yet.

They know that it feels bad when they can’t stop playing a game. They know that it feels weird when an app seems to know what they want before they do. They know that it doesn’t feel right that their stuff is actually not on their device and that they don’t know how it’s being used.

And their questions (the ones they kept coming back to) were exactly the right ones: Why don’t we have a choice? Why is it allowed to be this way? Who decided this was okay?

I didn’t have perfect answers, but what I did tell them was: These questions that you are asking… these are the questions that matter. Keep asking them.

Our public officials need this same lesson.

I didn’t talk to the kids about social media or specific platforms. We talked about algorithms, business models, and data—foundational problems that drive the behavior of digital products ranging from Instagram and TikTok to chatbots.

I worry that while the degree to which social media and digital platforms have warped childhood is starting to sink for people, everything that we’ve seen on social media—everything we are worried about—is happening all over AI. Just at a seismically larger and faster scale.

We need to stop focusing on specific products and platforms and focus on the roots of the problem:

• Algorithmic mediation that invisibly shapes our experiences

• Business models that feed off of “engagement” (aka addiction)

• Data that is extracted, amassed, and weaponized against users—jacking up prices, targeting kids with predatory content, and keeping you scrolling.

TikTok or Instagram or Roblox or whatever… the products are not THE problem. They are the SYMPTOM of the problem(s). If 8-year olds can get it, surely our policymakers can.