- Written statement of MIRI CEO Malo Bourgon to the AI Insight Forum (December 6, 2023)
- Ability to solve long-horizon tasks correlates with wanting things in the behaviorist sense (November 24, 2023)
- Thoughts on the AI Safety Summit company policy requests and responses (October 31, 2023)
- AI as a science, and three obstacles to alignment strategies (October 30, 2023)
- The basic reasons I expect AGI ruin (April 21, 2023)
- Misgeneralization as a misnomer (April 10, 2023)
- Pausing AI Developments Isn’t Enough. We Need to Shut it All Down (April 7, 2023)
- Truth and Advantage: Response to a draft of “AI safety seems hard to measure” (March 22, 2023)
- Deep Deceptiveness (March 21, 2023)
- Yudkowsky on AGI risk on the Bankless podcast (March 14, 2023)
- Comments on OpenAI’s "Planning for AGI and beyond" (March 14, 2023)
- Focus on the places where you feel shocked everyone’s dropping the ball (February 3, 2023)
- What I mean by “alignment is in large part about making cognition aimable at all” (February 2, 2023)
- A central AI alignment problem: capabilities generalization, and the sharp left turn (July 4, 2022)
- AGI Ruin: A List of Lethalities (June 10, 2022)
- Six Dimensions of Operational Adequacy in AGI Projects (June 7, 2022)
- Shah and Yudkowsky on alignment failures (March 2, 2022)
- Ngo and Yudkowsky on scientific reasoning and pivotal acts (March 1, 2022)
- Christiano and Yudkowsky on AI predictions and human intelligence (March 1, 2022)
- Ngo’s view on alignment difficulty (December 14, 2021)
- Conversation on technology forecasting and gradualism (December 9, 2021)
- More Christiano, Cotra, and Yudkowsky on AI progress (December 6, 2021)
- Shulman and Yudkowsky on AI progress (December 4, 2021)
- Biology-Inspired AGI Timelines: The Trick That Never Works (December 3, 2021)
- Soares, Tallinn, and Yudkowsky discuss AGI cognition (November 29, 2021)
- Christiano, Cotra, and Yudkowsky on AI progress (November 25, 2021)
- Yudkowsky and Christiano discuss “Takeoff Speeds” (November 22, 2021)
- Ngo and Yudkowsky on AI capability gains (November 18, 2021)
- Ngo and Yudkowsky on alignment difficulty (November 15, 2021)
- Discussion with Eliezer Yudkowsky on AGI interventions (November 11, 2021)
- Saving Time (May 18, 2021)
- Thoughts on Human Models (February 22, 2019)
- Embedded Curiosities (November 8, 2018)
- Subsystem Alignment (November 6, 2018)
- Robust Delegation (November 4, 2018)
- Embedded World-Models (November 2, 2018)
- Decision Theory (October 31, 2018)
- Embedded Agents (October 29, 2018)
- The Rocket Alignment Problem (October 3, 2018)
- Challenges to Christiano’s capability amplification proposal (May 19, 2018)
- A reply to Francois Chollet on intelligence explosion (December 6, 2017)
- Security Mindset and the Logistic Success Curve (November 26, 2017)
- Security Mindset and Ordinary Paranoia (November 25, 2017)
- AlphaGo Zero and the Foom Debate (October 20, 2017)
- There’s No Fire Alarm for Artificial General Intelligence (October 13, 2017)
- Ensuring smarter-than-human intelligence has a positive outcome (April 12, 2017)
- Using machine learning to address AI risk (February 28, 2017)
- Response to Cegłowski on superintelligence (January 13, 2017)
- AI Alignment: Why It’s Hard, and Where to Start (December 28, 2016)
- Safety engineering, target selection, and alignment theory (December 31, 2015)
- The need to scale MIRI’s methods (December 23, 2015)
- AI and Effective Altruism (August 28, 2015)
- Powerful planners, not sentient software (August 18, 2015)
- What Sets MIRI Apart? (August 14, 2015)
- Assessing our past and potential impact (August 10, 2015)
- When AI Accelerates AI (August 3, 2015)
- MIRI’s Approach (July 27, 2015)
- Four Background Claims (July 24, 2015)
- Davis on AI capability and motivation (February 6, 2015)
- Brooks and Searle on AI volition and timelines (January 8, 2015)
- Three misconceptions in Edge.org’s conversation on “The Myth of AI” (November 18, 2014)
- The Financial Times story on MIRI (October 31, 2014)
- AGI outcomes and civilizational competence (October 16, 2014)
- Groundwork for AGI safety engineering (August 4, 2014)
- Exponential and non-exponential trends in information technology (May 12, 2014)
- The world’s distribution of computation (initial findings) (February 28, 2014)
- Robust Cooperation: A Case Study in Friendly AI Research (February 1, 2014)
- How Big is the Field of Artificial Intelligence? (initial findings) (January 28, 2014)
- From Philosophy to Math to Engineering (November 4, 2013)
- Russell and Norvig on Friendly AI (October 19, 2013)
- Richard Posner on AI Dangers (October 18, 2013)
- Mathematical Proofs Improve But Don’t Guarantee Security, Safety, and Friendliness (October 3, 2013)
- How well will policy-makers handle AGI? (initial findings) (September 12, 2013)
- How effectively can we plan for future decades? (initial findings) (September 4, 2013)
- Transparency in Safety-Critical Systems (August 25, 2013)
- What is AGI? (August 11, 2013)
- AI Risk and the Security Mindset (July 31, 2013)
- What is Intelligence? (June 19, 2013)
- Friendly AI Research as Effective Altruism (June 5, 2013)
- When Will AI Be Created? (May 15, 2013)
- Five theses, two lemmas, and a couple of strategic implications (May 5, 2013)
- AGI Impact Experts and Friendly AI Experts (May 1, 2013)
- Once again, a reporter thinks our positions are the opposite of what they are (November 26, 2012)
- Three Major Singularity Schools (September 30, 2007)
- The Power of Intelligence (July 10, 2007)
