BI 151 Steve Byrnes: Brain-like AGI Safety

1

Introduction

00:00 • 3min

2

What's the Biggest Problem of Artificial Intelligence?

02:34 • 3min

3

How to Run Brain-Like Algorithms on Computer Chips

05:39 • 5min

4

Are You Accepted by Your Peers?

10:33 • 2min

5

Is There a Vision of Transformative AI?

12:37 • 2min

6

How Did You Get Into Artificial Intelligence?

14:24 • 2min

7

AGI Safety and AGI Alignment

15:56 • 3min

8

Agi

18:51 • 3min

9

Is AGI Going to Be Important in the Future?

22:03 • 2min

10

The Core Hypothesis Behind AGI Safety Concerns Is the Quote Orthogonality Thesis

23:58 • 4min

11

AGI Safety - Why Study Neuroscience?

28:02 • 3min

12

AGI Development via Reverse Engineering

31:02 • 2min

13

Learning From Scratch in the Brain

32:58 • 2min

14

Is the Brain Learning From Scratch?

35:22 • 2min

15

Reinforcement Learning and Steering Subsystems

36:55 • 2min

16

Is There a Difference Between Neural Architecture and Learning Subsystems?

39:24 • 2min

17

The Learning Subsystem and the Steering Subsystem in the AGI

41:19 • 4min

18

The Future AGI Systems Will Be Conscious

44:50 • 5min

19

Learning From Scratch

49:51 • 3min

20

Are You Trying to Get Safety Without Alignment?

52:27 • 4min

21

Is the AGI Trying to Escape the Box?

56:05 • 3min

22

How Would We Know if an AGI Is Conscious?

58:53 • 2min

23

Are Good and Bad Choices a Consequence of Fear of Death?

01:00:29 • 4min

24

Enemies Are Dependent on Context

01:04:52 • 3min

25

The Perfect Age to Model an AGI

01:07:53 • 2min

26

Are Retard Learning Algorithms Experiencing Pain?

01:10:01 • 2min

27

What Is Air Sets Interpretability?

01:11:33 • 4min

28

The Most Important Open Problems in AGI Safety

01:15:30 • 3min

29

Machine Learning

01:18:10 • 2min

30

AGI Safety Is Not an Issue

01:19:51 • 3min

31

How Many Years Ago Will We Have AGI?

01:22:30 • 5min

32

Brain Like AGI

01:27:52 • 3min