artificial intelligence: how it may become stronger and the risks involved.
This blog is mainly concerned with AI safety and the control problem; how humanity can make sure the outcome of developing strong AI will be beneficial (short answer: we don't know yet). In the post history below you will find articles about AI risk, approaches to safety research and the control problem, as well as more technical topics about state of the art methods in AI.
Inverse Reinforcement Learning Tutorial I (13/02/2018)
Risks of Artificial Intelligence (05/07/2017)
(planned) Introduction to AI Safety Research
(planned) The Free Energy Principle - Explaining how the brain works
(planned) Reparameterization Trick in Probabilistic Graphical Models
(planned) Reinforcement Learning - An Introduction
(planned) Deep Learning - How close is it to biological brains?
(planned) What is Intelligence?
Subscribe to Newsletter
Subscribe here to automatically receive new posts on thinking wires via email:
About This Blog
This blog was created by Johannes Heidecke, a graduate Student of artificial intelligence in Barcelona, Spain.