The Ethics Lab
A 501(c)(3) making AI legible to the people who'll inherit it — free curriculum, workshops, and INKY, an AI assistant built for schools.
Computer Vision
Neuroscience
Role
AI & Robotics Researcher
Platform
Free curriculum · workshops · INKY (AI for schools) · EthicsLabs.AI
team
Founder / director
platform
Web
The Real Problem
The Ethics Lab exists to make AI legible to the people who'll inherit it — kids. We build free curriculum, run workshops, and make INKY, an AI assistant for schools. Making that material is the part everyone worries about, and it's honestly the easy part. The hard part is everything after: getting it in front of kids at all.
Here's what makes it harder: kids don't go searching for AI-literacy lessons. They live on social feeds, where a careful explainer competes with everything else fighting for their attention. A workshop reaches the room it's held in. A curriculum reaches the classrooms that choose to adopt it. Neither of those is where most kids actually spend their time.
The pain points were consistent:
'Making the lesson is the easy part — getting a teenager to stop scrolling long enough to watch it is the whole game.'
'A workshop is a great afternoon, but it ends the moment the room empties out.'
'The kids who need this most are exactly the ones who'll never sign up for a course on it.'
The underlying issue wasn't the material—it was that making something and getting it seen are two different jobs. The Lab was built to make AI legible, and it does. Reaching the audience that matters most is a separate problem that good curriculum alone doesn't solve.
Finding the Fix
As founder and director, I run program design, partnerships, and direction — so I started by mapping how a lesson actually reaches a student: from the thing we make, through whoever puts it in front of a kid, to whether one ever sees it. Three problems stood out:
Reach isn't ours to control. Curriculum only lands where a school adopts it, and a workshop only reaches the room we're standing in.
The audience lives elsewhere. The kids we most want to reach are on social feeds, not signing up for lessons.
Attention is the bottleneck. Social feeds are organized around what grabs attention, not around what's good for someone — so a thoughtful explainer gets buried.
The teachers and parents I talked to described what they wanted as 'AI their kids can actually reason about'—what it is, what it can't do, where to be skeptical. But none of that matters if the message never reaches the kid in the first place.
So I focused on three fixes:
Meet kids where they already are — build for social feeds first, not just the classroom, so the material shows up in the places they actually spend time.
Make it feel native, not like homework. Short, honest, and worth watching on its own, so it earns attention instead of demanding it.
Lean on partnerships and INKY. Work with schools and put an assistant built for them directly in front of students, so reach doesn't depend on any single channel.
All doable with the team and budget of a small nonprofit. No big infrastructure required.
What Actually Happened
I tested the approach with teachers and a few students before scaling it — the people who actually see how kids engage. That's where I caught my mistakes.
For example: I'd assumed a good explainer would spread on its own if it was clear enough. But clarity and shareability are different things—kids share what feels like theirs, not what feels like a lesson. Making it clear wasn't enough; it had to be worth passing along.
The format evolved too. My first cuts were basically classroom lessons trimmed down. Kids scrolled right past them. I rebuilt them to open the way things that actually hold a feed open do — a hook first, the substance underneath. Small change in framing, big change in who stuck around.
We started with a small run first. The signal was quick—the pieces built for the feed held attention in a way the classroom cuts never had. We adjusted based on what landed, then widened it out.
What Changed
The clearest change was where attention went. Material built for the feed reached kids the classroom versions never would have, and treating distribution as the real problem — not an afterthought — is what moved things. The exact numbers are still early, but the direction isn't in doubt.
But my favorite feedback was qualitative
'This is the first time something about AI actually held my kid's attention.'
'My students found it on their own feed before I ever assigned it.'
'It doesn't feel like a lecture, it feels like something worth sharing.'
Kids passing it along without being told to is the metric I actually care about.
What I Had to Work With
The attention economy isn't on our side. We're competing for kids' attention against platforms engineered to hold it, with none of their budget. Every piece has to earn its place in the feed on its own merits.
A nonprofit's resources. A 501(c)(3) runs lean. No paid distribution muscle, no big media team—reach has to come from the work being good and from partnerships, not from spend.
A moving target. What kids watch and where they watch it shifts constantly. Anything built for one platform's format today may need rethinking tomorrow.
These constraints shaped everything. Instead of trying to out-spend platforms we can't out-spend, I focused on the few things a small nonprofit can actually win on: quality, trust, and partnerships.
What I'd Do Differently
I'd treat distribution as a first-class problem from day one, not something to figure out once the curriculum existed. Building great material first and worrying about reach later cost time — the reach problem is the actual problem, and it deserved attention earlier.
I'd also invest in partnerships sooner. Schools and educators are the most durable way to reach kids at scale, and leaning on that earlier would have compounded faster than chasing individual pieces of content.
What I Learned
Distribution is the product. I wanted to believe that if the material was good enough it would find its audience. It doesn't. Making the content is the easy part; getting it in front of kids is the work, and treating it that way changed everything.
Meet people where they are. What made sense to me (a clear, thorough lesson) wasn't what made sense to a kid mid-scroll. Building for how the audience actually behaves saved me from making something correct but unwatched.
Trust is the real advantage. A nonprofit can't out-spend the feed, but it can be the source kids and teachers actually trust on AI. That trust is slower to build and worth more than any single viral moment.
The Ethics Lab is a 501(c)(3) nonprofit. Learn more at EthicsLabs.AI.