When I say I work in education in SF, people react like I have the latest strain of the flu. They instantly back away. Most walk away entirely.
I get it. “Those who don’t do, teach.” And if you aren’t building in SF, you’re behind, at least in most social contexts.
Even at Worldwide Studios, we treat education as a means to an end. Accelerated learning is our mission, but learning alone isn’t the point. At our 24hr AI for Science hardware hack, the experimental data that comes out of end to end workflows. It’s pushing new tools as far as they can go building with cool people.
We don’t call these “education” events. They’re hackathons for professionals. And yet, everything we do is quietly redefining what education looks like in the age of AI.
So this weekend, we figured: we’ve done so much under every other label, why not throw up the e-word and see what happens?
We hosted an open build event. Five teams. Three hours. One question: what role does AI have in the future of education?
Even after hosting 100+ events, the wildest part is still the new normal for how much can get built in a single session.
Teams shipped deployment-ready iOS apps, agent orchestration layers wiring together voice, learning insights, and review agents using the latest tools like Rork, Anthropic, ElevenLabs and more.
And so with a few hours and the latest tools, here’s what 5 teams built exploring what AI native education could look like.
1. CareerGO for your first job
PokemonGO for young people looking for their first job or internship. Point your phone camera at any location and get a job pathway, salary breakdown, and information on skills needed. Work on quests to get experiences.
h/t Paul G. Erfan Al-Hossami Anyi Z. Sacha Lellouche and team. Built with Rork Max.
2. AI Native Soc Class for 1,000 students
Teacher side: upload documents and have them stored and indexed in vectorDB. Generate quizzes and classroom artifacts with the uploaded materials by class or section.
Student side: chat with the information uploaded with citations. Think customized NotebookLM.
Teacher side: review and assess all student assignments in a dashboard to make sense of personal learning journeys.
h/t Arlene W. Miraj Shah Oleksandra Pariy
Built with Anthropic Claude Code for SOC100H1 at the University of Toronto.
3. PromptLingo: a prompt tutor for mastering LLMs
Slop in, slop out. The quality of your prompts determines the quality of your output. There’s no easy way to do deliberate practice. This is a series of structured challenges for specific outputs using LLMs as a judge for getting prompts ready to paste into chat apps. Keep up to date with the latest models for best practices.
h/t Michael Raspuzzi. Yes, I tagged myself. I built this in Rork
4. Git Review: Understand a person by their projects
Skip the manual code review. Paste a GitHub profile and get an instant breakdown of someone’s actual skillset based on their repos.
h/t Bunyasit Fang and team
5. AI Readiness Voice Agent
Talk through your project with a voice agent that gives scoped, specific feedback and helps you figure out your next steps.
h/t Andrew Quintana Josh Richardson Ivana Pignatelli
Some personal takeaways
The tools aren’t the bottleneck. Institutional inertia is. A personalized learning dashboard for a 1,000-person class sounds like a glimpse of the future. But that was technically possible in 2023. When schools gate keep access to tools and move slowly on builder culture, they slow down the pace of innovation for everyone. Imagine if the first week of every class, students and teachers had to build their own solutions to the problems they actually face.
Stay current with the latest tools, but you don’t need to try all of them. If you need to build a mobile iOS app, Replit, Claude Code, and Rork are all options right now. You don’t need to use each one. But you can’t afford to not use any of them.
We need more solutions outside of traditional school. There’s a real gap between what AI engineers in industry and across startups are doing and what educators in the room see as useful. That gap shouldn’t be as wide as it is.
Here’s to more builds that bring engineers and educators together to bridge that gap.
Big thank you to Nate Padgett and the Studio 45 team for a community space enabling these kinds of community builds. Hi, I’m Michael, founder of Worldwide Studios, where we help people build with the latest in AI.
Reach out if you want to join a future build or collaborate.