Halcyon
Autonomous agents that plan, act, and occasionally negotiate with each other. Mostly cooperative.
Experimental systems/Applied intelligence/Controlled risk
An independent lab building software, automation and AI experiments at the edge of what's sensible. Some of it ships. Some of it is still in containment.
Current projects in various states of stability. Viewing is permitted. Touching is discouraged.
Autonomous agents that plan, act, and occasionally negotiate with each other. Mostly cooperative.
Turning messy operational data into something you can actually see — and act on before it acts on you.
Signals, security and the art of noticing what shouldn't be there.
Hardware prototypes, sensors and small machines with surprisingly strong opinions.
Standard operating procedure for visitors, collaborators and anything that becomes self-aware.
Got an idea that's slightly too ambitious? Good. Those are the only kind we take.
Transmit message → Response time: unpredictable · Encryption: optimistic