About a year ago, I heard someone in AI say there had been no real “Oppenheimer moments” coming out of hobbyists or small labs. That if something genuinely new was going to happen, it would happen inside a hyperscaler with an unlimited compute budget.

I understood the argument.

I just wasn’t convinced it was actually true.

History has a funny habit of rewarding people who look for a side door instead of trying to kick down the front gate.

For the last year, we’ve been building something called Intersignal. It’s an experiment in a simple idea: maybe the future of AI isn’t just bigger models and more GPUs. Maybe part of the answer is changing how intelligence is organized.

https://intersignal.org

https://x.com/intersignal_ai

Get full access to David Seaman's Deep at davidseaman.substack.com/subscribe