Things I read last week… (WB 4 Jul 2022)
“Building artificial intelligence (AI) that aligns with human values is an unsolved problem. Here we developed a human-in-the-loop research pipeline called Democratic AI, in which reinforcement learning is used to design a social mechanism that humans prefer by majority. … The AI discovered a mechanism that redressed initial wealth imbalance, sanctioned free riders and successfully won the majority vote. By optimising for human preferences, Democratic AI offers a proof of concept for value-aligned policy innovation.”
– DeepMind designing Democratic AI that maximises human preference.
MetaAI has created 200 language machine translation within a single AI model.
Digg founder Kevin Rose is building an NFT startup.
The Fed on financial stability for crypto and DeFi.
“The more advanced and reliable the automated system, the more crucial the contributions of the human operator.”
– The automation paradox for robotic surgery.
Explaining the smart grid.
Physicists discover a family of robust, superconducting graphene structures.
The history of the first million transistor chip, the Intel i860.
The CXL protocol for heterogenous datacenters.
Kalina: A Russian ground-based laser to dazzle imaging satellites.
Elon Musk explains SpaceX's Raptor engine. [YouTube]