Via productiva. Audio version.Listen now (16 min) | TL;DR: Introspection on how I do things and which rules and heuristics help me to be productive. Framed as Taleb's via negativa…
TL;DR: I let my friend Ava (who actually knows a thing or two about art!) experiment with DALL-E 2 for a bit. She allowed me to share her reflections…
and
TL;DR: deep reflections on names and identity, life-changing decisions, and mental renovations. And all of that in ~500 words!
6
2
Inferring utility functions from locally non-transitive preferences. Audio version.Listen now (14 min) | TL;DR: Fanboying JvN, then a nuts-and-bolts description of the von-Neumann-Morgenstern theorem. A connection to reward modeling…
Task Decomposition And Scientific Inquiry. Audio version.Listen now (11 min) | TL;DR: A curious asymmetry between making and criticizing, the scientific method as an approach to task decomposition, and a…
TL;DR: If you're a student of cognitive science or neuroscience and are wondering whether it can make sense to work in AI Safety, this guide is for you…
and
1
6
Puberty as Cause X? Audio version.Listen now | TL;DR: GiveWell-esque analysis of adolescents' suffering. Life satisfaction during puberty, ITN model, developmental neuroscience of the…
Previously in this series: Cognitive Biases in Large Language Models, Drug addicts and deceptively aligned agents - a comparative analysis, Compute…
The Unreasonable Feasibility Of Playing Chess Under The Influence. Audio version.Listen now | TL;DR: The wonderful tradition of playing chess drunk, Marr's levels of analysis, AlphaZero, and Iterated Amplification and Distillation.
TL;DR: Shameless advertisement for a paper some colleagues and I wrote. But also some pretty pictures of brain development, and some first principle…
1
1
Serendipitous connections: applying explanations from AI to the brain. Audio version.Listen now | TL;DR: A small shift in perspective (Elhage et al., 2021) helps interpret the ventral stream in the biological brain as the residual stream…
TL;DR: Inspired by Zillow's recent snafu I dig into the mathematics of adversarial attacks and recap some extreme value theory and optimal control. TW…
1