Four reasons I am excited about tech in learning

And I am not just writing this because I'm at ASU+GSV, the mother of all edtech conferences!

Photo: Alexey Dmidov for Unsplash.

I started my journalism career covering finance and economics, which means I spent an inordinate amount of my 20s and 30s thinking about incentives, markets and trade-offs.

Part of the reason I worry about tech is that the incentives are 100% not in kids’ favor.

Commercial tech—places like Meta and OpenAI—want engagement with their products. They don’t care whether our kids have balanced, tech-appropriate childhoods. They aren’t super worried if kids don’t develop the muscles to struggle with hard questions, or learn to write, or do boring math problems (a fairly essential part of learning math). We humans don’t like to think because thinking is hard, to paraphrase psychologist Daniel Willingham. Now every kid has a tool that helps them avoid uncomfortable thinking and instead get masterful answers in a fraction of the time! Mastery feels good when you are a novice, which every kid is.

So: the incentives for building the best AI tutor or AI-fueled feedback tool are winning market share, erecting barriers to entry, and trillion dollar market caps. The incentives for protecting kids from becoming cognitive vegetables or falling in love with bots rather than humans are…not quite so strong.

But I also live in the real world and I want to prepare my kids for it. AI is a part of that. According to a recent Microsoft and LinkedIn survey, 66% of leaders said they would not hire someone without AI literacy skills. In 2024 hiring for AI technical talent grew 30% faster than overall talent, according to another Microsoft survey (with the acknowledgment that all AI companies are incentivised to talk up its importance.) There’s a lot of evidence that the general job market for entry level jobs is hella bad (unless you’re a nurse, says the WSJ). I want my kids (and all kids) to use AI, understand it, and experiment with it, but also form judgments about its utility and its dangers.

I want to protect them from harm but prepare them for the future.

So why am I excited about tech in certain areas of education?

1. We’re equipped to do it better this time around

We were all kind of asleep at the switch with social media. I remember a Facebook presentation at Davos in the early 2010s when company executives (including Sheryl Sandberg) promised a more connected world that would naturally be kinder, more understanding and more democratic. That DID NOT happen and social media has played a starring role in driving us apart with algorithms that reward hate, misinformation, social comparison and industrial level, 24/7 envy (thanks guys!)

So where’s the good news?

We are not asleep at the switch anymore. We are getting better at understanding risks and opportunities. We are organizing, using the courts and finding nuance in use. We get that dosage matters (see my Atlantic article), and that kids cannot be on screens in school or out of school all the time. We are starting to realize (or, possibly, to remember) how essential human connection is to human flourishing, and that classrooms are complex social places where that needs to happen. We can hopefully see what tech is good for (high school math graphene) and what it’s not good for (tranquilizing kids for hours on end).

Hopefully we all push like hell to make sure kids get a better balance at home and at school: tech when it helps, and a healthy dose of human when that’s what we are building for. That means figuring out the risks tech poses but also leaning into the opportunities for progress. Which leads me to…

2. Education is pretty broken and it does need fixing

A quarter of kids in the US are chronically absent. A ton of young people find school boring, and believe that it will not prepare them for the real world. Many technologies have promised to drag education out of the dark ages. Their track record isn’t great.

My hope is that the existential threat of AI will force us to look way more closely at what happens in schools, including what needs to be rethought (this taxonomy from Burning Glass is interesting). AI will be part of this. This of course comes with a massive caveat, which is that we need to protect what’s essential to learning. Struggle (or friction). Discomfort. Collective effervescence (read Dylan Kane talking about it; it’s awesome). Trust in classrooms.

3. We can move from over-promising to precision learning

I loved Robin Lake’s post about precision learning. Personalised learning has largely been a bust. But precision learning, like precision medicine, means identifying a problem (like a disease), understanding its pathology, and treating that. We know kids are different and need different support. AI can provide that.

“AI gives educators the potential to understand, diagnose and respond to students’ learning needs with a specificity that was previously impractical at scale,” Lake writes. “It can rapidly surface a child’s learning gaps and strengths in math and recommend targeted interventions. But that information alone does little; AI’s power lies in being embedded in professional workflows, guiding adults toward specific, evidence-based actions and tracking whether those measures improve learning over time. To effect genuine change, AI must be accompanied by a reevaluation of the systems that contain it.”

We cannot continue educating to the average. We can tailor to the individual while also developing kids’ social responsibilities and muscles as members of a society.

We have let tech distract us and divide us. Now we have to do the work of rebuilding the muscles of real life (see: Center for Teen Flourishing), connecting (The Rithm project) and reclaiming our attention (a10d) We need classrooms that are social spaces where rigorous learning takes place, and grade inflation is eliminated. This is hard, but worthwhile.

4. Some evidence suggests kids can use AI well

Some early research from OpenAI (potential bias noted) suggests that kids who are trained on AIs that make them work harder (more Socratic questioning and less “here’s the answer”) chose that over general purpose LLMs. It’s probably too early to draw massive conclusions from this but what I take from it is:

• Plenty of kids want rigor (stay with me)

• But only if think they can master it

• Most don’t think they can. And so they check out: better to not try and fail than try and fail

• But if stretched, and supported, kids will choose to go back for more because doing hard things feels better than doing easy things.

I am no dummy. I get who is holding all the cards here. Big Tech still has limited incentives to do what’s right for kids. But legal settlements are starting to rain down. Social norms are starting to change. We are waking up to the shitshow we’ve helped create (or failed to protect against). Anger is useful. It can mobilize. It’s better than resignation.

And we aren’t the only ones who are angry. Gallup’s most recent data shows anger about AI has increased nine points since 2025, to 31%, among 14-to-29-year-olds (Gen Z). Gen Zers’ agreement or strong agreement that they feel excited about AI has dropped 14 percentage points to 22%, while hopefulness has fallen nine points to 18%. The report underscores an incredibly complex set of emotions young people are carrying about AI, namely that they have to use it to get ahead, but they fear doing so will make them dumber.

“Even as 56% of Gen Zers acknowledge that AI tools can help them complete their work faster, they see a cost: 8 in 10 Gen Zers (80%) believe that relying on AI to complete tasks faster will likely make learning more difficult in the future.”

My mama bear anger may be big, but their Gen Z anger is terrifying to Big Tech. That’s their prime hunting ground to land lifetime users. This young generation is learning they have agency; though I hate that they also have to carry so many emotions about tech. As if climate change and polarization and student debt wasn’t enough, now we have AI anxiety.

One of my all time favorite songs is American Tune by Paul Simon:

I don’t know a soul who’s not been battered

I don’t know a friend who feels at ease

I don’t a know dream not that’s not been shattered

Or driven to its knees.

I get that’s not exactly a light ending but here’s my drift: life is hard, but we’re in this together. We don’t have to solve everything, just commit to trying to do a little better every day. Marginal gains, they’re called in economics. Not as enticing as silver bullets, but definitely more effective.

I’ll soon be publishing my thoughts on the fascinating ASU+GSV conference itself.

---

Recommendations

Gaia Bernstein on Whether this is Big Tech’s Tobacco Moment.

Jacqueline Nesi of Techno Sapiens on why it may not be.

Derek Thompson on whether the Smartphone Theory of Everything is right.

My piece in The Atlantic on what happened when a teacher banned Chromebooks from his classroom.

And a book: The Correspondent by Virginia Evans. It will break your heart in 100 pieces and restore your hope in humanity. All in a novel told through letters.