Current papers are at the top. Every version is archived on Zenodo, and later versions correct earlier ones in the open.
Why "artificial" is a structurally harmful label and "super" the wrong replacement, the alignment mechanism as a testable hypothesis, and a three-tier framework for standing that grows with capability.
Coming shortly on Zenodo · request an early copy →A framework for intrinsically aligned digital superintelligence (DSI), with machine-checked foundations: 26 results in Rocq (Coq), every assumption listed. Earlier versions' claims about formal verification and timelines are corrected in its errata.
v1.3.1 coming shortly on Zenodo · request an early copy →An analysis of internal Meta documents reported by Reuters on revenue from scam advertising.
Read the analysis →