​

The kids are all right

Most of the time, I consider how AI is changing work for those of us who are already adults. However, as the parent of three teenagers, I'm under no illusions: the real decisions about how AI will shape the workplace will be made by their generation. In ten years, our software, policies, and practices will be defined by the norms that teenagers are building today.

So for this week's case study, I wanted to get a view into how teenagers think about and use AI in school.

Who I considered

I surveyed 200 active middle and high school students and asked them six simple questions about how they use AI in school. The group was half middle schoolers and half high schoolers. I included a diverse group of genders, races, and school types: public and private, large and small, urban and not.

In the interest of naming my biases: Everyone I included is based in the greater Seattle, WA area. That means the patterns I saw may not be nationally representative (yet). However, I was particularly interested in the experiences of this narrow group, because Seattle is a tech-forward market where students are likely to have high familiarity with AI.

The tl;dr:

• Most of the kids are already actively using AI in school
• Most of the kids are using AI as a reasonable learning aid, not to plagiarize content wholesale
• Middle school students are already pushing the envelope on AI use faster than high school students

AI use is already pervasive, despite school rules

Over the last year, we've seen a massive effort to set AI policy in every area of public life: workplace, government, law, finance, healthcare, and of course education. You have to feel for leaders trying to set these policies. The AI landscape is changing so rapidly that defining reasonable policy is nearly impossible work.

But while institutions are setting policies, people aren't waiting around. They're putting the technology to practical use. Students are no exception.

80% of the kids in the survey have used AI to help with homework. The kids in this Seattle-based data set are likely to adopt AI faster than the national norm. But based on these numbers, it's reasonable to believe that at least half of teens nationally have used AI in similar ways.

But kids aren't using AI the way you might think: Less than a third of the kids are turning in documents, images, or videos that are completely AI-generated. Rather, teens' main use cases are practical and mostly assistive. For instance, they're using AI to come up with topic ideas, write outlines, or check math or coding problems.

Kids also aren't easily impressed by AI output. On the whole, they think AI output is about as good as what they do on their own, not better. However, they do see AI as making it a lot faster to achieve an acceptable standard of work. One kid left a comment that summed it up: "I don't use AI to get brilliant work. I use it to get good-enough work, really fast."

Sorry high schoolers, you're already old

So far, we've looked at questions where middle school and high school students responded the same way. But in several areas, middle school students were way more AI-forward. I don't know why this surprised me, because I see this play out in my own household. But I found it really striking in the survey.

Middle schoolers are four times more likely to disagree with their school's rules about AI use. 21% of the middle schoolers disagree with the rules, compared with just 5% of the high schoolers.

That said, whether or not they agree, most of the kids follow their school's guidelines about AI use. Fewer than 25% of the kids in both age groups have used AI without their teachers knowing.

But even here, middle schoolers are pushing the envelope more. Twice as many middle schoolers say they've used AI for an assignment without their teacher being aware.

But by far the most fascinating insight from this whole mini-study is this: Whatever school policy dictates, middle schoolers just don't see using AI as cheating the same way high schoolers do.

Kids these days! (Are adults tomorrow.)

Look, because of how I chose them, my survey group is likely to be at the very far edge of AI adoption. But the rate of AI use among this cohort is stunning, and it points to how things are likely to go more broadly in the next few years.

Love the trends or hate them, this generation will bring their AI experiences to the workplace over the next few years. In 15 years, they'll be managers, and in 25 years, they'll be running companies. Their POV about the best use cases for AI and the ethics around those use cases will win.

What do you think?

Thanks for reading!

Kieran

​Catch up on nerd processor case studies | Subscribe​

kieran@nerdprocessor.com
​Unsubscribe · Preferences​