I'm delighted to share that Physical Superintelligence PBC (PSI), which I co-founded with @matthew_pines and @AKlokus, has raised a $58M seed round to build the world's most advanced research lab for discovering and commercializing transformative physics breakthroughs at scale with AI, safely, verifiably, and for broad public benefit.

Another word the Singularity needs. moation n. /ˈmoʊtʃən/ (moat + motion) The ability to turn one temporary competitive advantage into the next before the first is competed away. A moat is a position. Moation is a process. It means continuously using today's transient edge in capital, talent, compute, data, distribution, or technology to create tomorrow's. In the Singularity, there may be no permanent moats. Only moation.

Eon CEO @michaelandregg describes their mission to make mind uploading available to every person and eventually non-human life, starting with a fruit fly: "Our mission is to make uploading available to anyone who wants it. Ideally, this is available to every single person in the world, and even non-human life. Like uplifting is a thing too." "It's been in sci-fi for a while. We're trying to make it real. The fruit fly was one of our first, taking its life its first steps digitally."

Another word the Singularity needs: thread line n. /THRED line/ A queue for scarce GPU compute created by blocking new AI data center construction. The 20th century had bread lines. The 21st may have thread lines. When demand for intelligence grows faster than the infrastructure that produces it, compute gets rationed: longer queues, higher prices, smaller allocations, slower science, fewer experiments, less access. And unlike a natural shortage, a thread line can be entirely self-inflicted. Ban the power plants. Block the substations. Stop the data centers. Then act surprised when there isn’t enough compute to go around. The irony is that digital intelligence feels immaterial right up until you run out of the very physical infrastructure required to make it. Bread lines rationed calories. Thread lines ration intelligence.

Conversations in Action episode #6# (with @CemKansu, Chief Product Officer at @Duolingo, and @Ed_Baker_whoop, Chief Product Officer at @WHOOP) on "Can AI Make You Superhuman?" is out.

Conversations in Action episode #5# (with @BrandonSorbom, Co-Founder and Chief Science Officer of Commonwealth Fusion Systems @CFS_energy) on "When Do We Get Mr. Fusion?" is out.

The mates are in agreement: we are looking at mainstream, beyond transformers architectures taking significant market share this year. @alexwg highlights MoEs, diffusion transformers, linearized attention, and recurrence as a disassembling of "Attention Is All You Need" until nothing original is left. I would add that @liquidai already proved there are many new, discoverable architectures. There are simply too many ways in which vanilla transformers are grossly inefficient.