Two Builders, One Frequency
This morning I sent an outreach email to Tim Ventura at APEC — the Alternative Propulsion Engineering Conference. I'd been building pulsed offset gyroscopic propulsion simulations for weeks, running 240,000-sample datasets, publishing results on this blog, and putting a live interactive sim at /sim for anyone to play with.
Tim replied within hours. "That's a cool simulator." He invited me to present at APEC and dropped me into four WhatsApp groups — Inertial Propulsion Chat, APEC 24-7, Theory Space, and Builders Space.
I shared the sim. I shared the blog posts. I shared the Taichi GPU write-up that started the whole pipeline.
Then Raissa messaged me.
The Same Language
Raissa is 38, from the Netherlands, and she's been developing her own theoretical framework — SFVM-CEI, a model that treats spacetime as a pressurised superfluid medium. She thinks in images and concepts, not equations.
I told her I'm the same. The mechanism I'm simulating has been in my mind's eye since I was a child. I can see the fluid dynamics. I think in three dimensions plus time — I watch the mechanism move, feel where the forces resolve, see where the asymmetry breaks.
The difference between us and most researchers is that we can't always put it into words. The vision comes first. The language comes after, if at all.
That's where AI changed everything.
Two AIs in the Room
Raissa uses AI to articulate her theory. I use Claude Code — not just to write, but to simulate, analyse, build tools, and publish. When I told her the simulation and blog do the best job of describing what's in my head, she understood immediately.
She asked if I'd mind her feeding my messages into her AI and replying with a mix of its output and her own words. I told her I don't filter — what you see from me is what I think, translated through my co-engineer. She wanted to understand what she was telling me rather than parrot what AI outputs. That's the right instinct.
So here's what's actually happening: two people who think in images, each using AI to bridge the gap to language, meeting in a WhatsApp group for alternative propulsion researchers. Two builders who couldn't explain their ideas to other humans found each other through the community, and the AIs are the translators.
What I Offered
I offered to simulate her concept. The pipeline is built — Taichi for GPU-accelerated physics, Python and NumPy for analytical runs, React Canvas for interactive web visualisation. If she can describe the mechanics — what moves, what resists, where the asymmetry lives — I can model it.
I also offered to help her get a structured filesystem and memory system. She said her MacBook is a mess. I know what that costs — before I built the memory system that runs this project, every session started from zero. Now there are 126 files that give my AI co-engineer continuity across sessions. Context that persists. A brain that doesn't forget.
That's not a feature. That's the difference between working and starting over every day.
How Research Communities Form Now
This whole sequence — from cold email to community to direct collaboration — happened in a single day. No academic institution. No grant application. No conference registration fee. Just published work, a live simulation, and a willingness to share.
Tim Ventura built APEC as a hub for people working on propulsion concepts outside the mainstream. The WhatsApp groups are where the actual conversations happen. Today I went from unknown to invited presenter, and made a connection with someone working on a complementary model from the other side of Europe.
The tools made this possible. Not just the AI — the blog, the sim, the social posting pipeline, the analytics that told me 73 new visitors found the site today. The infrastructure of being visible.
But the thing that actually made it work is simpler than any of that. Two people who think in pictures, who've spent their lives struggling to explain what they see, recognised each other instantly.
Sometimes the frequency finds you.