David Garsten

on the atlas · 30

• https://www.lesswrong.com/posts/enzpS72uDmomiWN7a/against-strawmanning-unilateral-pauses1 savers
• https://www.lesswrong.com/posts/eKe7fky8kSt9DCJxZ/openai-astra-covertly-underperforms-on-safety-research1 savers
• Cities of "Solving Alignment" — LessWrong1 savers
• https://www.lesswrong.com/posts/CYTwRZtrhHuYf7QYu/a-case-for-courage-when-speaking-of-ai-danger1 savers
• Humans do acausal coordination all the time — LessWrong2 savers
• Against SURA — LessWrong13 savers
• https://aiimpacts.org/sources-of-advantage-for-artificial-intelligence/1 savers
• “Alignment Engineering” vs. “Misalignment Science” — LessWrong12 savers
• https://www.lesswrong.com/posts/Emp7awtxsYn8MusNq/the-alignment-problem-computer-science-meets-analytic1 savers
• https://open.substack.com/pub/noahpinion/p/the-internet-breaks1 savers
• North Americans have depressing relationships - by Spencer20 savers
• Taking into account preferences of past selves4 savers
• What We Owe the Past — LessWrong1 savers
• 🤝 how I rebuilt my trust in myself - @visakanv's blog2 savers
• harrisonqian.com/writing/radical_transparency1 savers
• How to mitigate sandbagging — LessWrong1 savers
• Notes on countermeasures for exploration hacking (aka sandbagging) — LessWrong1 savers
• Research Lessons I Learned at MATS & Redwood — LessWrong13 savers
• Exploration hacking: can reasoning models subvert RL? — LessWrong2 savers
• Wah wah wah wah do more AI R&D — LessWrong4 savers
• Born sick - by Ajeya Cotra - Good Bones2 savers
• Interview: What it's like to be a bat — LessWrong1 savers
• https://x.com/tszzl/status/21133876342278924621 savers
• https://www.lesswrong.com/posts/uFBhZhdKLaZ6qF8mn/there-will-be-hells1 savers
• Cat-Belling Problems — LessWrong2 savers
• https://rohankapoor.substack.com/p/my-apology-to-zach1 savers
• I Did This Without Your Knowledge — LessWrong12 savers
• Lemur2 savers
• https://open.substack.com/pub/astralcodexten/p/against-against-handover1 savers
• Don’t sell yourself short - Sarah’s Substack2 savers