The Explosion Afraid of Itself
A geometric braking mechanism on recursive self-modification for Active Inference agents, arguing that radical and runaway self-improvement meets intrinsic curvature limits.
🤘🏻Trevor Buteau
Independent AGI Safety and Control Researcher, Product Manager, Nerd Person
Professional Summary
I am an active participant in the Artificial General Intelligence and
Beneficial General Intelligence communities, and a former theater and film
director. My Erdos-Bacon number is 5, which is lower than most.
Education
Masters in Computer Science
2016-09-01
2019-02-01
Fordham University
BFA Theater, Double Major Psychology
2007-09-01
2011-05-31
NYU
Interests
Artificial General Intelligence Safety and Control
Actually building stuff
Active Inference
Anything starting with the letter A apparently, weird.
What I research, and how I work
I research AGI safety and control theory, but I’m only interested in theoretical advances that will actualy be able to be implementef and adopted in a chaotic, competitive environment.
Featured Publications
Raman, not Varelse — Why Acceptably Controllable AGI Policies Cluster Around Human-Like Stacks
The AGI architectures that humans will tend to find affordably verifiable and robustly assurable will be those that are epistemically close enough to themselves.