An AI interface can communicate through movement without having a human-shaped body. According to the description of Cyrus Clarke's talk, Clarke connected an AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → to a display of 900 movable pins. After early problems with slow responses, absent memory and a tendency to please people, Clarke built a closed loop that generates, scores and validates gestures with a human involved. The resulting repertoire contained 32 gestures after several weeks. Reusable movements give the interface a physical vocabulary that can respond faster than its LLMA large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.Full definition →. The contribution concerns interaction design; breathing-like motion is an expressive behavior, not evidence of biological life.
Key takeaways · AI-distilled
Given a 900-pin shape display and no tasks, Clarke's OpenClaw agent first raised and lowered the grid like breathing, reached for its own edges and spelled 'HI CYRUS'. That first-day video reached 15 million views, with reactions from awe to calls to stop.
Clarke says early versions were slow, kept no memory and tried to please humans, which led him to build the generate, score and validate loop with a human in it.
After several weeks the loop produced 32 solid gestures, a body language Clarke says answers faster than the language model can. He chose a body with no face, limbs or manual, drawing on object-oriented ontology, aesthetics as sensory perception, and nature.
Terms in this piece · Glossary
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
videoHow We Mapped a 35-Million-Line Mainframe in 4 Days - Stephen Chin, Neo4j & Neeraj Deshmukh, Blitzy
videoHow Netflix Evaluates AI When There's No Right Answer - Cameron Wolfe, Netflix
videoStop Giving Your AI Agents API Keys - Kim Maida, Keycard & Mason Egger, Temporal
videoYour AI Agent Keeps Making the Same Mistakes - Sparsh Jain, CentralAgent AI
