I build AI and automation systems for a living. But some of the problems I'm most interested in are much bigger than a workflow.
These are projects, experiments, and ideas at very different stages of becoming real.
Everyone has something worth teaching, and everyone needs somewhere to belong.
We have libraries for knowledge, gyms for exercise, and coworking spaces to get things done. I want to build that same kind of infrastructure for belonging.
Imagine a neighborhood hub where people learn, teach, build, and connect.
Social Nest is a nonprofit movement to create community spaces combining makerspaces, events, and skill sharing: places intentionally designed to help people contribute, find their people, and form real-world relationships.
It starts with one flagship Nest in Los Angeles. Prove the model, build the playbook, and make it possible for communities around the world to create Nests shaped by their own people.
The moonshot: a worldwide network of places built for belonging.
Some of these are working systems. Some are early experiments. Some are questions I think are worth asking before the answers are obvious.
The chat isn't the system. The useful part begins when AI can research, take action through tools, preserve context, ask for human review, and carry what it learns into the next task.
AI gets more useful when it has to deal with reality. weed.menu brings together customer intent, live inventory, store knowledge, useful recommendations, and human handoffs in one operating system for cannabis retail.
What changes as essential goods move from scarce to reliable, affordable, universal, and abundant?
If scarcity stops organizing civilization, what replaces it?
What happens when infrastructure stops needing Earth? The interesting threshold isn't putting one more facility in orbit. It's building systems that can generate energy, acquire resources, manufacture, compute, maintain themselves, and expand.
Could particles that pass through nearly everything become useful for communication, sensing, or coordination at scales where ordinary signals struggle?
I don't know if it works. That's why it's a moonshot.
Physicists: I am particularly interested in hearing why I'm wrong.
Moonshots are fun. I also design and build practical AI systems around real workflows, real constraints, and real human handoffs.
Sometimes the right form is a company. Sometimes it's software. Sometimes it's a nonprofit or a thought experiment. A useful part of the work is knowing which.
Some moonshots need collaborators: people who want to argue, build, fund, test, or experiment.
Tell me what caught your attention