HTML Is All Agents Need

Official Schedule Context

LLMs are great at writing code. So the question we kept asking was: can they write code that

produces a video? We thought it would be easy. The reality was a year of trying. We started with

massive prompts to get very mediocre output. We made it more agentic to iterate and improve its

output. This worked okay but wasn't production-ready. Eventually we tried Remotion. It got us

deterministic video, but the React framework kept boxing the agent in. The more guardrails we added,

the safer and more boring the outputs got. When we utilized plain HTML, CSS, and JavaScript, the

creativity came back to the output. So we set out to build a video rendering framework on top of

HTML. But it needed to work with Gemini Flash. Why? Because one tell that a framework is fighting an

agent is needing the biggest model just to get usable output. So from there we shaped the framework

around what small models could reliably author. That left one real engineering question: can we keep

the freedom of HTML and still render a deterministic MP4? Browsers don't want to do that. Image

decoders, font loaders, and animation clocks all run async on their own schedule. Great for

performance. Terrible for "render the same pixels every time." Throughout, we iterated constantly

with agentic loops and self-improving evals to test out the framework, find issues in our renderer,

and shape a set of skills that gave the agents Taste instead of guardrails. This talk is what it

took to get there.

No official session recording transcript was found by exact title match on the AI Engineer YouTube channel during this run.

Pending transcript synthesis when an official recording or confirmed matching video is available.