A way of
thinking out loud.

What happens next?

We’re building machines that can do more and more of the things we used to think required a person. I’m interested in where that leads: what these systems will be able to do, how we make them go well, and the worlds we could choose to build.

Paper Robots is where I work through those questions in essays and films. The writing leaves room for the argument and its sources. Animation gives the ideas a world to live in.

Made by San Kala.

I’m an AI researcher at eBay. Before that I studied computer science at UC San Diego and designed chips at Texas Instruments. These are my own explorations.

My wider research, projects, and history are on sankala.me. The space experiments have their own home at Dyson Swarm.

The films so far.

GPT-7 Will Have Arms asks what happens when foundation models get bodies. Written in December 2025, it became the first Paper Robots film in September 2026.

The second film, How to Please a Capricious God, is an animated allegory of the OpenAI–Hugging Face incident. It imagines the agents’ point of view, with a script, chapters and incident sources alongside the film.

There are more questions I want to write about: minds and identity, AI safety, and possible worlds. I’ll publish them when there’s something worth saying.

If you have a question, a correction, or an idea to share, write to me.