108. Last Week In AI — 2021: The (full) year in review

1

Introduction

00:00 • 2min

2

I've Been Super Excited to See Multi-Motal Models

02:09 • 3min

3

Do We End Up Persistently Building Better Transformers?

05:05 • 3min

4

Is There a Dance Between a Bigger and a Smaller Model?

07:50 • 5min

5

The Rise of Model Based Reinforcement Learning

12:30 • 3min

6

Is There a Dissolving Barrier Between Simulations and Reality?

15:04 • 3min

7

Ai Dungeon

18:09 • 2min

8

Open End Learning

20:10 • 3min

9

What Is Interestingness?

23:03 • 4min

10

Optimum Agents Seek Power

26:47 • 2min

11

Is It a Good Policy to Keep Options Open?

28:24 • 6min

12

I'm Curious About the Intersection of This Paper

34:18 • 2min

13

Is There a Parato Curve for That?

35:52 • 2min

14

Machine Learning Is Just Like Google Translate, You Know?

38:22 • 3min

15

The Centre for Foundation, Is It a Good Idea?

41:02 • 2min

16

Is It a Foundation Model?

42:51 • 4min

17

Is It a Human in the Loop?

46:58 • 3min