Nitya Kumar was 22 when she landed a product design role at Meta straight out of art school. It was the kind of offer new graduates dream about. But by 2024, two years in, she could see the writing on the wall: the entire industry was pivoting toward AI, and her skill set was still rooted in traditional product design.
She knew she needed to learn AI. But instead of diving into formal courses or creating a strict study plan, she took a different approach — one that would eventually become her signature philosophy.
"I stopped seeing my career as a ladder with a focus on moving upward," she explains. "In my view, all work and no play keeps the Big Tech offers away, so I started treating my career like a science lab, giving myself the freedom to experiment."
Coming from an art school background with no formal machine learning training, Kumar taught herself through YouTube videos and friends. She set up a simple but effective routine with an accountability buddy: one hour of Cursor (the AI code editor) per day, seven days a week, equals one functional AI prototype.
Her first project was something playful and personal — a game that could detect and track the gestures of a user dancing in their room. Kumar loves dance, and building something connected to her passion made the learning process feel less like work and more like play. It helped her develop what she calls a "vibe coding" workflow — using different AI tools for different parts of the process, from Gemini to Cursor to Figma MCP.
But she didn't stop at safe, portfolio-friendly projects. She unleashed what she calls her "inner mad scientist." She built a tool that generated matcha recipes, using Gemini to develop prompts she'd feed into Claude to vibe-code the product. She built weird, fun things that had nothing to do with traditional design portfolios.
These quirky experiments, she believes, are what made her stand out in interviews. Instead of generic case studies, she'd show interviewers her matcha recipe generator or her dance gesture game. They'd laugh, they'd debate matcha flavors, and they'd play around with the tools she'd built. It was memorable. It was human. And it demonstrated real AI fluency in a way that resume bullet points never could.
After roughly four months of intentional AI experimentation, Kumar landed a UX design job at Adobe working on agentic AI experiences — including conversational AI assistants. She started in November 2025.
"Treating my career like a science lab has not only helped me transition into an AI role without a machine learning background, but it's also helped me succeed at it," she says. "I feel like I've mastered vibe coding, fueled my creativity, and become a better designer."
Today, every other Friday, she leads an AI playground workshop for other designers at Adobe in India. They come together online to experiment with new tools and share how they're using AI-assisted workflows. Designers come to her for support with debugging or building prototypes, and her confidence grows as she helps them.
Her advice to other designers considering the AI shift? Stop waiting until you feel "ready." The technology is evolving so fast that everyone is learning as they go. Build weird things. Have fun with it. And don't underestimate the power of a good side project — even (or especially) the silly ones.