The render economics of AI-generated video made synthetic human likenesses a studio-budget proposition for most of the 2010s. Victor Riparbelli founded Synthesia in 2017 on the thesis that the compute floor would drop far enough to put that capability inside a laptop workflow. Four years of middling results followed before he changed course.
Where the cost curve meets the founding idea
Riparbelli's original ambition was specific. He wanted Synthesia to let any individual produce Hollywood-quality video without a crew or a camera. The compute existed to do this in 2017, but at a price that served studios, not individuals. That gap between technical feasibility and economic feasibility is where Synthesia spent its first four years.
Riparbelli has described those years as producing middling results. The company was building but not breaking through.
The shift, and the investor it attracted
After four years, Riparbelli made a major strategic change. Mark Cuban invested in the company at some point in that arc. In Riparbelli's telling, getting Cuban on board was the easy part of building Synthesia. That's a pointed framing. It places Cuban's check among the simpler problems, which means the harder ones were elsewhere.
Why the timing matters
ChatGPT reached consumers in late 2022. By then, Riparbelli had been working inside the AI video thesis for five years, from Synthesia's 2017 founding. He wasn't repositioning into generative AI after the term went mainstream. He was already there, absorbing the years when the market didn't yet know what to do with what he was building.
Riparbelli is 34.