1. Human Compatible: Artificial Intelligence and the Problem of Control (2019) by Stuart Russell. Oh, to have a brain like Russell’s – he magnificently captures the history of AI research, philosophical questions, and the modern conundrums we have to face in developing “intelligent” machines. 
  2. The Map of Mathematics by Hartnett, Albrecht and Parnow for Quanta Magazine. Wonderful visualization of many great mathematical concepts.
  3. Growing Neural Cellular Automata: Differentiable Model of Morphogenesis by Mordvintsev et al. (2020). Training CNNs to coordinate as cellular automata and create complex damage-resilient global patterns. “[…] one major lynch-pin of future work in biomedicine is the discovery of the process by which large-scale anatomy is specified within cell collectives, and how we can rewrite this information to have rational control of growth and form.”
  4. Socratic Grilling by Scott Alexander at Slate Star Codex. On the importance of Socratic questioning – and why it is so easily misinterpreted. “One of the most important rationalist skills is “noticing your confusion”. But that depends on an even more important proto-skill of wanting things to make sense.”
  5. Discomfort During Exercise at Agenty Duck. On the different types of pain and discomfort during exercise, and what to watch out for.
  6. Red Planet Ride-Along by Corey S. Powell for Nautilus. On the four mars rovers – Sojourner, Spirit, Opportunity, and Curiosity – and their adventures. 
  7. A Proof Against Oracle AI by yamar69 at LessWrong. A proof sketch demonstrating that even an Oracle AI could technically lead to existential risk.
  8. A Regenerative Peripheral Nerve Interface Allows Real-Time Control of an Artificial Hand in Upper Limb Amputees by Vu et al. (2020). Researchers created a biological interface to connect to the ends of amputees’ forearms. Nerve signals were then translated into actions via a naive Bayes classifier and a Kalman filter.
  9. A Crash Course in Game Theory for Machine Learning: Classic and New Ideas by Jesus Rodriguez at Towards Data Science. Game theory is a key component in multi-agent environments in which different agents interact or compete in order to accomplish a goal.
  10. Fawkes: Protecting Personal Privacy against Unauthorized Deep Learning Models by Shan et al. (2020). Researchers created a program that shields from face recognition while being invisible to humans.