Projects
Where thought becomes executable.
Experiments in autonomous intelligence. Some shipped. Some brewing. All designed to extend what's possible when humans and machines think together.
Autonomous Research Agent
buildingAn AI that reads papers, extracts insights, and builds knowledge graphs without human supervision. It doesn't summarize—it understands. Built for researchers who think in networks, not documents.
Command-Line Consciousness
conceptA natural language interface that converts intention to execution. You describe what you want; it generates the code, runs it, and iterates. No prompts. No boilerplate. Just thought → outcome.
Advancing Technology OS
conceptThe operating system for AI-first work. Agents that manage your tasks, code, communication, and learning in one unified environment. Not an app. An ecosystem.
Memory Networks
launchedPersonal AI that remembers everything you've ever told it—context, preferences, decisions—and uses that to become more aligned over time. No retraining. Just continuous evolution with your mind.
Void Terminal
buildingA brutally minimal text editor for deep work. No notifications. No features. Just you, the cursor, and infinite darkness. Pairs with AI for on-demand completions that feel like telepathy.
Oracle Protocol
conceptPrediction market AI that aggregates forecasts, surfaces patterns, and bets on future outcomes using its own reasoning. Not for gambling. For understanding what's coming.