https://shash42.substack.com/archivedailyhttps://shash42.substack.com/aboutweeklyhttps://shash42.substack.com/p/what-does-your-benchmark-actually2026-06-27monthlyhttps://shash42.substack.com/p/does-generalization-generalize2025-12-27monthlyhttps://shash42.substack.com/p/how-to-game-the-metr-plot2025-12-22monthlyhttps://shash42.substack.com/p/the-research-process2025-12-22monthlyhttps://shash42.substack.com/p/automated-scientific-discovery-as2025-12-22monthlyhttps://shash42.substack.com/p/what-if-things-went-well2025-12-22monthlyhttps://shash42.substack.com/p/not-all-bits-are-made-equal2025-12-22monthlyhttps://shash42.substack.com/p/it-is-time-to-move-on-from-mcq-evaluations2025-12-22monthlyhttps://shash42.substack.com/p/counting-down-capabilities-to-agi2025-12-22monthlyhttps://shash42.substack.com/p/coming-soon2025-12-22monthly