Designing agent incentives to avoid side effects

Choosing a baseline

Choosing a deviation measure

Effects of the design choices

Future directions

--

--

--

We research and build safe AI systems that learn how to solve problems and advance scientific discovery for all. Explore our work: deepmind.com

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Using AI and Machine Learning to beat the competition

A Computer Could Become Your Doctor

Robot giving a red rose to a woman with dark hair

Facial Recognition Technology in Law Enforcement

Could AI End Partisan Gerrymandering?

Is Your Chatbot Smart Enough to Understand Your Customers?

Artificial Intelligence — The fear of humanity to lose control

Pros and Cons of Artificial intelligence

coursework help

Introducing Counteractual

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
DeepMind Safety Research

DeepMind Safety Research

We research and build safe AI systems that learn how to solve problems and advance scientific discovery for all. Explore our work: deepmind.com

More from Medium

DeepMind’s PoG Excels in Perfect and Imperfect Information Games, Advancing Research on General…

Diffusion models

[Olivia Reading Notes] “Decision Transformer: Reinforcement Learning via Sequence Modeling”

Review of Reinforcement Learning Papers #13