They Came to Learn AI. They Left Knowing Their Own Value.

On August 10, our interns walked into PhoenixTeam headquarters in Arlington, Virginia, for the Showcase Showdown. Each pod had ten minutes in front of a shark-tank-style panel to present and demo what it had built. For many of them, it was the first time they had met in person. One intern told us they had never seen online connections carry into real life like that.

The summer was a 70-work-day experiment. At the start of the summer, we brought 24 young people into our company. We gave them AI tools, real problems, and real responsibility. Then we asked them to do the one thing no AI tool can do for them: lead with their own judgment.

They delivered. When the summer ended, we asked everyone involved how it went. Here’s what they told us, what we learned, and why next summer will be even better.

The bet we made

AI is changing work faster than most classrooms and first jobs can keep up. Young people hear a lot about which tools to learn. They hear far less about what those tools can’t do, or about the value they bring themselves.

Human Futures is PhoenixTeam’s answer. It’s our youth learning and career-readiness initiative, built around one mission: prepare young people to work alongside AI, with human judgment first.

Our first summer internship put that mission to the test. Interns had two jobs. Day to day, they worked alongside a manager on one of our teams. They also joined a pod: a small team with a sponsor and a real business problem to solve. Weekly sessions covered how AI works, how businesses run, and how to communicate like a professional.

The bet was straightforward. Give young people real work and real trust. Teach them to use AI well. Then see what they do with it.

Real work for a real company

The pods didn’t get practice problems. They took on projects PhoenixTeam actually needed.

Nearly every intern said they were proud of what their pod produced. So did every single pod sponsor.

“The work produced was both useful to PhoenixTeam now, and aspirational for what we could do in the future.”

Confident with AI. Clear on what it can’t do.

Interns learned to build with AI, not just talk about it. One described using Claude Code, Replit, GitHub, and AI agents to develop applications, debug code, automate repetitive tasks, and prototype new ideas. Another said they now use AI as a productivity tool rather than a novelty.

But confidence with tools was never the whole goal. We wanted interns to understand what they bring that AI can’t. So we asked them.

Their top answer was creative and design judgment, followed by accountability and communication. As one intern explained, people design and build for other people, and that takes empathy, not just pattern recognition. Another put it even more simply:

“While AI is good at building things, it is not very good at making decisions.”

That’s the instinct we set out to build. Not fear of AI. Not blind faith in it. A clear sense of where people need to lead.

Growing as people, not just workers

The skills mattered. The growth mattered more.

Of everything we measured in Year One, that’s the result we’re proudest of. Managers saw the growth in everyday moments. One described an intern who absorbed what he had learned and acted on it to unblock his team, without being called in. Another was impressed by what an intern built with Claude to help evaluate HR tools.

Interns felt that investment, too. Every intern said they felt supported by their manager.

Our interns became teachers

One pod carried the mission even further. From August 3 to 7, the Human Futures pod ran AI Builder Week for 12 kids ages 10 to 16. The goal was a first real experience with AI: learn how it works, learn how to prompt, and build something by the end of the week.

The week was hands-on from start to finish. But the most important learning happened away from the screen.

That ratio says a lot about how we approach AI. The tools matter. The thinking behind them matters more. Watching our interns teach that idea to kids was one of the best parts of the summer.

AI Builder Week · August 3 to 7

Built for Year Two

It was a great summer. It wasn’t perfect. The same honest feedback that showed us what worked also showed us what to do better, and every lesson is already shaping what comes next. Here’s what the Class of 2027 can look forward to:

• Sharper projects from day one. Every pod starts with a defined need, clear expectations, and a known finish line.

• More time together. Many of our interns worked remotely, and the Showcase showed what happens when people finally share a room. Next year, we’ll build that bond early and often, not save it for the last week.

• More hands-on learning. More hands-on challenges and real case studies, balanced for business-minded and technical interns alike.

• A smoother start. One central hub for resources and updates, and an easier first week.

• Wider doors. Our partners pushed us to think bigger. We’re recruiting beyond our own networks, matching interns to roles that fit their strengths, and inviting other companies to sponsor interns and projects.

To the Class of 2027

We asked this year’s interns what advice they’d give next summer’s class on day one. Two answers stood out.

When we asked interns whether they’d recommend Human Futures, most gave it a perfect 10. One summed up the summer in words we’ll carry into Year Two:

“This program is a true life changing experience, and something that is so valuable for this generation.”

Thank you

Year One happened because a lot of people said yes to something new. Thank you to Tela Mathias for the vision, and to PhoenixTeam’s leadership for backing Human Futures from the start. Thank you to the managers and pod sponsors who gave their time. And thank you to our interns, who showed up, took risks, and told us the truth.