Brain Inspired

BI 151 Steve Byrnes: Brain-like AGI Safety

Oct 30, 2022

01:31:17

Creator website

Episode guests

Steve Byrnes

Highlights

AI Chapters

Episode notes

Learning From Scratch in the Brain

02:24

AGI Development via Reverse Engineering

01:56

How Would We Know if an AGI Is Conscious?

01:35

Introduction

3min

What's the Biggest Problem of Artificial Intelligence?

3min

How to Run Brain-Like Algorithms on Computer Chips

5min

Are You Accepted by Your Peers?

2min

Is There a Vision of Transformative AI?

2min

How Did You Get Into Artificial Intelligence?

2min

AGI Safety and AGI Alignment

3min

Agi

3min

Is AGI Going to Be Important in the Future?

2min

10.

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

4min

11.

AGI Safety - Why Study Neuroscience?

3min

12.

AGI Development via Reverse Engineering

2min

13.

Learning From Scratch in the Brain

2min

14.

Is the Brain Learning From Scratch?

2min

15.

Reinforcement Learning and Steering Subsystems

2min

16.

Is There a Difference Between Neural Architecture and Learning Subsystems?

2min

17.

The Learning Subsystem and the Steering Subsystem in the AGI

4min

18.

The Future AGI Systems Will Be Conscious

5min

19.

Learning From Scratch

3min

20.

Are You Trying to Get Safety Without Alignment?

4min

21.

Is the AGI Trying to Escape the Box?

3min

22.

How Would We Know if an AGI Is Conscious?

2min

23.

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

4min

24.

Enemies Are Dependent on Context

3min

25.

The Perfect Age to Model an AGI

2min

26.

Are Retard Learning Algorithms Experiencing Pain?

2min

27.

What Is Air Sets Interpretability?

4min

28.

The Most Important Open Problems in AGI Safety

3min

29.

Machine Learning

2min

30.

AGI Safety Is Not an Issue

3min

31.

How Many Years Ago Will We Have AGI?

5min

32.

Brain Like AGI

3min

Support the show to get full episodes, full archive, and join the Discord community.

Steve Byrnes is a physicist turned AGI safety researcher. He's concerned that when we create AGI, whenever and however that might happen, we run the risk of creating it in a less than perfectly safe way. AGI safety (AGI not doing something bad) is a wide net that encompasses AGI alignment (AGI doing what we want it to do). We discuss a host of ideas Steve writes about in his Intro to Brain-Like-AGI Safety blog series, which uses what he has learned about brains to address how we might safely make AGI.

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.