Essays
- Newest
- Popular
- Notable
- Statistics
- Meta-Science
- Decision Theory
- Order Statistics
- Crypto/Bitcoin
- Dark Net Markets
- AI: Safety
- Generative AI: Image
- Generative AI: Fiction
- Deep Learning
- Computer Science
- Haskell
- Cognition
- Psychology
- Behavior Genetics
- Economics
- Economics: Tech
- Domestic Cats
- Practical
- Design
- QS: Sleep
- QS
- Politics
- Epistemology
- Ethics
- Philosophy
- Literary Criticism
- Anime
- Fiction: Prose
- Fiction: Verse
- Docs
- Docs: Science
- Docs: Crypto
- Docs: NGE
- Docs: Anime
- Wikipedia
- Reviews
- Personal
This is the website of Gwern Branwen. I write about AI, psychology, & statistics. I am best known for my writings about AI scaling, poetry & anime neural networks, darknet markets & Bitcoin, blinded self-experiments, and dual n-back & spaced repetition.
For information about my site’s philosophy & method, see the About page; for the website features & implementation, see the Design page; for information about myself, my use of other websites, and contact information, see the about-me page; for information about new pages, see the Changelog ( = new), or new links. For an annotated version of this site index page, see the long index.
To use dark-mode () or reader-mode (), or disable popups (), or search the site (), use the floating toggle bar in the lower-rightupper-right corner (); for more, see the help (?) page.
Newest
Popular
Notable
Statistics
Meta-Science
Decision Theory
Order Statistics
Crypto/Bitcoin
Dark Net Markets
AI: Safety
Generative AI: Image
Generative AI: Fiction
Deep Learning
Computer Science
Haskell
Cognition
Psychology
Behavior Genetics
Economics
Economics: Tech
Domestic Cats
Practical
Design
QS: Sleep
QS
Politics
Epistemology
Ethics
Philosophy
Literary Criticism
Anime
Fiction: Prose
Fiction: Verse
Docs
Docs: Science
-
Radiance (Carter Scholz)
-
“You And Your Research” (Hamming)
-
“Sanity & Survival” (Hofstadter)
-
“Climb Mount Improbable” (Dawkins)
-
“The Sword Of Achilles” (Asimov)
Docs: Crypto
Docs: NGE
Docs: Anime
-
“Otaku Talk” (Okada et al)