magfrump . 9/16/26 magfrump . 9/16/26 Agency Again Read More magfrump . 9/1/26 magfrump . 9/1/26 Our Models Are (Sometimes) Better Than Our Benchmarks Read More magfrump . 8/25/26 magfrump . 8/25/26 My Questions About Agency Read More magfrump . 8/19/26 magfrump . 8/19/26 LLM-driven Code Review Process Read More magfrump . 6/30/26 magfrump . 6/30/26 The Problem with Chat Read More magfrump . 4/14/26 magfrump . 4/14/26 Reasons to be Worried About AI Read More magfrump . 4/9/26 magfrump . 4/9/26 Engaging vs. Engagement Read More magfrump . 4/7/26 magfrump . 4/7/26 All Alignment Research is Capabilities Research Read More magfrump . 4/7/26 magfrump . 4/7/26 Dirty dishes on the counter Read More magfrump . 3/27/26 magfrump . 3/27/26 The Church-Turing Anthithesis Read More magfrump . 3/25/26 magfrump . 3/25/26 Metaethics and Mathematical Constructivism Read More magfrump . 3/12/26 magfrump . 3/12/26 Why Do You Trust Your Compiler More Than Your Coworker? Read More magfrump . 10/10/25 magfrump . 10/10/25 Belegarth Video Analysis Read More magfrump . 7/9/25 magfrump . 7/9/25 MASK evals with small models Read More magfrump . 7/9/25 magfrump . 7/9/25 Sandbagging thought experiment Read More magfrump . 4/3/25 magfrump . 4/3/25 Bring out your thoughts Read More magfrump . 3/20/25 magfrump . 3/20/25 LLM Fact Checking Read More magfrump . 5/28/24 magfrump . 5/28/24 LLM Tokenizer Compression Read More magfrump . 5/21/24 magfrump . 5/21/24 Community Notes 2 Read More magfrump . 5/11/24 magfrump . 5/11/24 Fictional Governments Read More Older Posts
magfrump . 3/12/26 magfrump . 3/12/26 Why Do You Trust Your Compiler More Than Your Coworker? Read More