June / July / Aug 2025 Reading

Books

• 2666, Roberto Bolaño (2004)
• A Portrait of the Artist as a Young Man, James Joyce (1916)

Papers

• Cartridges, Eyuboglu et al. 2025
• The Era of Experience, Silver and Sutton, 2025
• Linear Layouts: Robust Code Generation of Efficient Tensor Computation Using F2, Zhou et al., 2025
• Emergent Misalignment, Wang et al., 2025

Blogs / Essays

• Good writing, Paul Graham
• Why I don't think AGI is right around the corner, Dwarkesh Patel
• To Think Deeply About Simple Things, Greg Burnham
• Dwarkesh Patel dinner party, Yuxi Liu
• What happened in the 2010s?, Rohit Krishnan
• Reinforcement learning and general intelligence, Finbarr Timbers
• Training AI to play Super Smash Bros. Melee, Eric Gu
• The Case for More Ambition, Jack Morris
• Specification Engineering, Joshua Purtell
• Notes on Friction, Rohit Krishnan
• The Gentle Singularity, Sam Altman
• Notes on Anthropic's How we built our multi-agent research system, Simon Willison
• Malleable software, Ink and Switch
• Q-learning is not yet scalable, Seohong Park
• Basic Facts about GPUs, Damek Davis
• Reflections on OpenAI, Calvin French-Owen
• On "ChatGPT Psychosis" and LLM Sycophancy, John David Pressman
• Vandermonde matrices are merely exponentially ill conditioned, Ethan Epperly
• Nihilism Is Not Enough, Peter Thiel