2025:

• 02-07: Your LLM Can Be Faster Than You (Can) Think

2024:

• 12-01: ‘Tlön, Uqbar, Orbis Tertius’: Fascist Idealism

• 11-29: Virtual comments: LLM idea for writing LessWrong posts

• 11-23: What do you do after ‘winning’ an AI arms race?

• 11-15: The State of Chinese AI

• 11-14: What do we mean by ‘diminishing returns’ in scaling?

• 09-27: Abs-E, or, speak only in positives § text2epositive.py test

• 08-24: An LLM acceptable-use policy: authors must improve samples

• 08-22: Hardware Hedging Against Scaling Regime Shifts

• 08-21: You should write more online—it’s still a good time

• 08-16: Marvel’s Silver Age of Comics: Humanizing Mythology

• 08-07: What is an ‘AI warning shot’?

• 07-30: Towards Better RSS Feeds for Gwern.net

• 07-29: A Gambler’s Verity World: Sampling-Without-Replacement

• 07-22: Flaws in Midjourney’s new personalization feature

• 06-19: Is OpenAI OK? How would we know & what would it look like?

• 06-16: GPT-3 Semantic Derealization

• 05-12: Why ‘Tools For Thought’ Work Fails: Lots of Work for Tools, Not Thought

2023:

• 11-12: Why do writers still underestimate LLMs?

• 09-30: Against Caring About Subtle Poisons

2020:

• 02-09: Crowdsourcing The Best GPT-2-1.5b Poetry

2019:

• 02-19: ThisWaifuDoesNotExist.net

2018:

• 10-20: WBE & DRL: a Middle Way of imitation learning from human brains

2011:

• 12-18: Inverse p-zombies: the other direction in the Hard Problem of Consciousness