A Peek at the Future of AI Governance
OpenAI and Google’s DeepMind recently published some really interesting research on a new approach to using human preferences in shaping machine learning algorithms.
In this piece, I dissect this research to try to make it easier to understand and then investigate some of its implications. The biggest, I think, has to do with how we humans might govern AI in the future. And that’s important stuff.
#MachineTraining #MachineLearning #Governance #AI
P.S. Yes, I shared an earlier version of this but made some tweaks and so am re-sharing.