Vitaly Bulatov

To start my time with @rerundotio, I want to give an idea as to what I'll be working on over the next few months. I included a small sneak peek of things using the new daggr framework from @Gradio, which is a node-based Gradio space orchestrator. I've been thinking about everything I've worked on so far and how much coding has changed. As the first major project, I'll be consolidating all my disparate repos into a single cohesive repo. The goal is to have a definitive end-to-end pipeline for egocentric and exocentric data collection and to explain how @rerundotio fits in at each stage: from raw sensor collection, prelabeled data ingestion, data quality, annotation, and finally model training and evaluation. BUT with the specific goal of trying to write as little code myself, I'll be heavily leaning on using Claude Code and Codex. I'm honestly not sure how this will go, but I'll be making this a public experiment to show what works and what doesn't. I think a critical part of this will be to use the right abstraction, some seed code, and the tooling I mention below to try to make this successful. The plan is to make small, modular, composable pieces that will ideally work well with agentic development. I'll be leaning on the following @rerundotio - deep visual debuggibility + rrd artifacts that agents can interface with pyserde - used for typesafe serialization and deserialization of Python dataclass with numpy and JSON. beartype + jaxtyping - runtime type checking generally, as well as numpy dtype and shape. Perfect for reducing agent errors @Gradio - API disguised as a UI. Perfect for agent use to validate inputs/outputs and for easy human-in-the-loop visual validation daggr - orchestration layer between Gradio nodes for end-to-end testing on a single timestep playwright cli - for agent visual debugging + automation and lots of testing infrastructure Anyway, I'll be sharing more over time as I continue to build in public!

Marc Andreessen on the untapped potential of AI agents: "You could have an AI bot that basically raises money to make a movie and then spends the money on image generation and sound generation." "Maybe even hiring actors... set designers or graphic artists or... sound effects people, musicians." "Or let's take a more serious case... you could have an AI bot that's doing protein folding and literally coming up with cures, doing personalized medicine for cancer patients." "You could easily imagine having an economic mechanism for that... hypothetically, you could have the equivalent of like a GoFundMe, but on the blockchain, for people to basically be able to pay an AI bot to cure their cancer." @pmarca