Oly on AI

Oly on AI

Home
Notes
AI Governance
Predicting the Shape of AI
My Research
Fun
Oly
Archive

Research

Cooperation and Alignment in Delegation Games
You need both! (and also fairness)
 • 
Oliver Sourbut
Deliberation Everywhere: Simple Examples
We pre­vi­ously dis­cussed a con­cep­tual al­gorith­mic break­down of some as­pects of goal-di­rected be­havi­our with the in­ten­tion of in­spiring…
 • 
Oliver Sourbut
Deliberation, Reactions, and Control
Tentative Definitions and a Restatement of Instrumental Convergence
 • 
Oliver Sourbut
Breaking Down Goal-Directed Behaviour
When we speak about en­tities ‘want­ing’ things, or hav­ing ‘goal-di­rected be­havi­our’, what do we mean?
 • 
Oliver Sourbut
Conditions for mathematical equivalence of Stochastic Gradient Descent and Natural Selection
Analo­gies are of­ten drawn be­tween nat­u­ral se­lec­tion and gra­di­ent de­scent (and other train­ing pro­ce­dures for para­met­ric al­gorithms). It…
 • 
Oliver Sourbut
Some real examples of gradient hacking
Spec­u­la­tion and an in­vi­ta­tion to make fur­ther sug­ges­tions or clarifications
 • 
Oliver Sourbut
© 2025 Oliver Sourbut
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture