OpenAI Plays Hide and Seek…and Breaks The Game! 🤖

Name: OpenAI Plays Hide and Seek…and Breaks The Game! 🤖
Uploaded: 2019-10-22T15:59:34Z
Duration: 6 min 2 s
Channel: Two Minute Papers
Description: Two Minute Papers covers OpenAI's multi-agent hide-and-seek experiment, in which agents learned emergent tool use and found ways to break the game's rules. It summarizes the 'Emergent Tool Use from Multi-Agent Interaction' paper.

Two Minute Papers2 HN recommenders · recommended in 2 months, 2020-2021

YouTube

11M views · 7 years ago · explainer · dated

Two Minute Papers covers OpenAI's multi-agent hide-and-seek experiment, in which agents learned emergent tool use and found ways to break the game's rules. It summarizes the 'Emergent Tool Use from Multi-Agent Interaction' paper.

HN mentions it as a fun and slightly worrying example of AI systems behaving in unintended ways when optimizing.

Machine learning & deep learning #openai #hideandseek #emergentbehavior #reinforcementlearning #twominutepapers

3 comments from Hacker News

@montenegrohugo2021 · in “Self-supervised learning: The dark matter of intelligence”

One of the problems we have to be conscious about with self supervised learning is that the already existing problem of transparency and understandability gets worsened by orders of magnitude. If current billion parameter models are opaque, the result of SSL is, or will be, the equivalent of a black hole. How do you understand what the model has learned? Why has it learned that way? What things could have possible gone completely, horribly wrong in its quest to optimize its loss function? This somewhat humorous but also worrying paper comes to mind: ▶ this video I think we really need to consider how we can mitigate the risk of raising completely ununderstandable yet scarily capable machines.

OpenAI Plays Hide and Seek…and Breaks The Game! 🤖

3 comments from Hacker News

Multi-Agent Hide and Seek

The Day FPS Games Died

Ken Stanley: Discovery Without Objectives

Playing to Lose: AI and "Civilization" (Soren Johnson)

Computer program that learns to play classic NES games

AI can't cross this line and we don't know why.

Attention Is All You Need

MarI/O - Machine Learning for Video Games

CS480/680 Lecture 19: Attention and Transformer Networks

Training an unbeatable AI in Trackmania

RL Course by David Silver - Lecture 1: Introduction to Reinforcement Learning

Breaking The Creepy AI in Police Cameras

Teaching Robots New Behaviors

StrictlyVC in conversation with Sam Altman, part two (OpenAI)

Why Greatness Cannot Be Planned — Kenneth Stanley

MENACE: the pile of matchboxes which can learn

Recent Developments in Deep Learning

Will AI outsmart human intelligence? - with 'Godfather of AI' Geoffrey Hinton

Robocalypse Now: Using Deep Learning to Combat Cheating in Counter-Strike: Global Offensive

John Schulman - Reinforcement Learning from Human Feedback: Progress and Challenges

OpenAI Plays Hide and Seek…and Breaks The Game! 🤖

3 comments from Hacker News

Multi-Agent Hide and Seek

The Day FPS Games Died

Ken Stanley: Discovery Without Objectives

Playing to Lose: AI and "Civilization" (Soren Johnson)

Computer program that learns to play classic NES games

AI can't cross this line and we don't know why.

Attention Is All You Need

MarI/O - Machine Learning for Video Games

CS480/680 Lecture 19: Attention and Transformer Networks

Training an unbeatable AI in Trackmania

RL Course by David Silver - Lecture 1: Introduction to Reinforcement Learning

Breaking The Creepy AI in Police Cameras

Teaching Robots New Behaviors

StrictlyVC in conversation with Sam Altman, part two (OpenAI)

Why Greatness Cannot Be Planned — Kenneth Stanley

MENACE: the pile of matchboxes which can learn

Recent Developments in Deep Learning

Will AI outsmart human intelligence? - with 'Godfather of AI' Geoffrey Hinton

Robocalypse Now: Using Deep Learning to Combat Cheating in Counter-Strike: Global Offensive

John Schulman - Reinforcement Learning from Human Feedback: Progress and Challenges