blogposts

Recent

1. 42

   We know how to train AI, but what should we train it for?

   Aug 23link
2. 41

   Reality distortion fields exist at every level

   Aug 19link
3. 40

   The information content of sensory experience

   Aug 12link
4. 39

   How Einstein used intuition to discover general relativity

   Aug 9link
5. 38

   More things should be named after bad people

   Aug 7link
6. 36

   Theoretical research will turn practical

   Aug 23link
7. 35

   A Seed Crystal (2/2)

   Aug 23link
8. 34

   A Seed Crystal (1/2)

   Aug 23link
9. 32

   Wittgenstein, Pirsig, and Feynman help me figure out what to write about

   Aug 23link
10. 31

    Model Uncertainty Without Priors (Part 2)

    Aug 23link
11. 30

    Model Uncertainty Without Priors (Part 1)

    Aug 23link
12. 29

    Simulating the Patriotic Gene

    Aug 23link
13. 28

    Truth-maxxing

    Aug 23link
14. 27

    If you don't know what to do, consider what you won't do

    Aug 23link
15. 26

    The inescapable bugs of the human operating system

    Aug 23link
16. 25

    For the next year, I'll be emailing you twice a week

    Aug 23link
17. 24

    Most people don't know how to make money

    Aug 23link
18. 23

    Optimism is useful and necessary

    Aug 23link
19. 22

    Active Learning in Autonomous Highway Merging

    Aug 23link
20. 21

    Research, Readings, and Revelations

    Aug 23link
21. 20

    Value in Contrastive Goal-Conditioned RL

    Aug 23link
22. 18

    Plasticity as Adaptation vs. Informational Influence

    Aug 23link
23. 17

    On Bayesian Neural Networks and Martingale Posteriors

    Aug 23link
24. 16

    Policy Gradient and the Generalized Advantage Estimator

    Aug 23link
25. 15

    On-Policy vs. Off-Policy Algorithms

    Aug 23link
26. 14

    Zen and AI Research

    Aug 23link
27. 13

    Reduce uncertainty only about the things you already know matter

    Aug 23link
28. 11

    Matched Attunement

    Aug 23link
29. 10

    On LLM Psychosis

    Aug 23link
30. 9

    Advice is Projection

    Aug 23link
31. 8

    The Problem with Holding Water

    Aug 23link
32. 7

    The Geometry of Knowledge and What I'm Betting On

    Aug 23link
33. 6

    In Defense of Cringe

    Aug 23link
34. 5

    Why is Dual Control Hard?

    Aug 23link
35. 4

    Gaussian Processes and Meta-Learning

    Aug 23link
36. 3

    You'll Never Finish and That's Good

    Aug 23link
37. 2

    New Austin

    Aug 23link
38. 1

    Fear and Love in Ann Arbor

    Aug 23link