Beren's Blog
Thoughts on (AI) consciousness
Millennials as the Forever Generation
The Limit of Prediction is not Omniscience
Most Algorithmic Progress is Data Progress
Do We Want Obedience or Alignment?
Should we be behaviourst about an AI's values?
Preliminary Thoughts on Reward Hacking
Why Not Sparse Hierarchical Graph Learning
The Scaling Laws Are In Our Stars, Not Ourselves
Current neural networks are not overparametrized
Maintaining Alignment during RSI as a Feedback Control Problem
Review of Mind Children By Hans Moravec (1988)
Intellectual Progress in 2024
A Retrospective on Active Inference
Right to Left (R2L) Integer Tokenization
Alignment likely generalizes further than capabilities.
The Unconditioned Distribution of Current Open LLMs
Capital Ownership Will Not Prevent Human Disempowerment
Integer tokenization is now much less insane
Alignment In The Age Of Synthetic Data