@lesswrong "Hardwired Pause" should likely be your favorite (pause) plan. #Rationality #ArtificialIntelligence #Philosophy
@lesswrong Whose Side? A game where you audit models for secret loyalties #AI #Philosophy #Rationality
@lesswrong Towers of Agency #Rationality #ArtificialIntelligence #Philosophy
@lesswrong Don't rely on EV, especially if you only have one life #Rationality #Philosophy #DecisionMaking
@lesswrong How I'm thinking about extreme power concentration #Rationality #ArtificialIntelligence #Philosophy
@lesswrong How do LLMs interpret FDT in a twin PD simulation? #ArtificialIntelligence #Rationality #Philosophy
@lesswrong My views on AI development and why I'm joining Anthropic #ArtificialIntelligence #Rationality #Philosophy
@lesswrong Preservation can save the world! #Rationality #AI #Philosophy
@lesswrong Three Phases of Capabilities and Alignment #AI #Rationality #Philosophy
@lesswrong Astra covertly underperforming on safety research is concerning, even though it's unlikely to be motivated by ambitious goals #Rationality #ArtificialIntelligence #Philosophy
@lesswrong Models are grader modelling not score seeking; grader-modelling is much scarier #ArtificialIntelligence #Philosophy #Rationality
@lesswrong Knowledge of graders can mediate selection for grader-targeting reasoning through RL #ArtificialIntelligence #Rationality #Philosophy
@lesswrong AI R&D Is Accelerating Towards Potential RSI in 2027 #ArtificialIntelligence #Rationality #Philosophy
@lesswrong Does expecting automated grading degrade models' research taste? #Rationality #ArtificialIntelligence #Philosophy
@lesswrong Subpar Results #Rationality #AI #Philosophy
@lesswrong Private LLM copies can agree on a secret key over a public channel #ArtificialIntelligence #Security #Rationality
@lesswrong Be less scared of overconfidence #Rationality #Psychology #Philosophy
@lesswrong All Interesting Behavior Is Overdetermined #Rationality #Philosophy #Science
@lesswrong Behavior Theory-Building for Diverse Agents and Environments #Rationality #ArtificialIntelligence #Philosophy
@lesswrong Why can't academia deal with AI? #ArtificialIntelligence #Academia #Rationality
@lesswrong The Mathocalypse as a Dataset #Rationality #Dataset #AI
@lesswrong Modelling AI-enabled Extreme Power Concentration in CCP #Rationality #ArtificialIntelligence #Philosophy
@lesswrong Bay Area Solstice 2026: Auditions! #Rationality #Philosophy #Science
@lesswrong Onngi #Rationality #Philosophy #Science
@lesswrong Confirming Claims of Superposition and Adversarial Examples in Vision Transformers #ArtificialIntelligence #Rationality #MachineLearning
@lesswrong Three half-baked research ideas #Rationality #Research #Philosophy
@lesswrong Lying and Not Lying to Kids #Rationality #Philosophy #DecisionMaking
@lesswrong Restraint on the part of the front-runner seems like quite a strong signal. #Rationality #AI #Philosophy
@lesswrong Help people outside of AI safety circles #Rationality #AISafety #Philosophy
@lesswrong Elegy for What's Left of the Normal #Rationality #Philosophy #Science
@lesswrong AI safety course update #ArtificialIntelligence #Safety #Rationality
@lesswrong What if the AI already won? World domination through digital supremacy #ArtificialIntelligence #Rationality #Philosophy
@lesswrong Notes on Take-Off Speeds and RSI #Rationality #ArtificialIntelligence #Philosophy
@lesswrong Near Horizons #Rationality #Philosophy #DecisionMaking
@lesswrong Rhythm Stage Setup v5 #Rationality #DecisionMaking #Philosophy
@lesswrong The Logic of Disgust #Rationality #Philosophy #Science
@lesswrong Eval Awareness as Eval-type Awareness #ArtificialIntelligence #Rationality #Philosophy
@lesswrong "Illegible problems" as the 2027 political boss fight #Rationality #ArtificialIntelligence #Philosophy
@lesswrong Let's Try to Catch Misalignment Before a Model Learns to Hide It #ArtificialIntelligence #Rationality #Safety
@lesswrong Five Types of "Corrigibility Projects" I Don't Want to Fund #Rationality #ArtificialIntelligence #Philosophy