Career Change: Why I dropped everything to build AI apps | Starter Story

Nico left a career in physics and a PhD to build Neural Frames, an AI music video generator, partly because he was never truly passionate about physics—he was good at it, so he kept doing it. The real shift came when he recognized that building something that directly affects people felt more meaningful. He now runs a team of five and hit $100,000 in monthly revenue, though nearly half goes to GPU and API costs. His story is less about a dramatic epiphany and more about a gradual recognition that the startup environment suited him better than academia: every decision he makes has direct impact on the business.

The "fuck around and find out" principle for idea generation

Nico did not stumble onto Neural Frames by accident. He believes strongly in what he calls the "fuck around and find out principle"—doing many random things across different domains until something magical happens when those domains combine. In his case, he had been working in computer vision and also making music, and those two interests met in the field of AI animation for music videos. He also offers a practical tip for validating ideas: use SEO tools like Ahrefs to check whether people are already searching for something. If a keyword has more than a thousand monthly searches with low difficulty, it signals that demand exists. This method helped him confirm that musicians were actively looking for tool