The Gradient: Perspectives on AI

Jeremie Harris: Realistic Alignment and AI Policy

56 snips
Jun 29, 2023
Ask episode
Chapters
Transcript
Episode notes
1
Introduction
00:00 • 2min
2
How I Got Into AI
01:34 • 4min
3
The Physicist to AI Person Pipeline
05:10 • 3min
4
The Importance of Transcendent Reality
08:21 • 3min
5
The Limits of Sufficient Powerful Models
11:47 • 2min
6
The Challenges of Transitioning Fully to Being an AI Person
13:20 • 4min
7
How to Avoid Basic Mistakes When Building a Startup
16:51 • 3min
8
The Importance of Discrimination in AI
19:36 • 3min
9
The Paradigm Shift: How AI Is a Dangerously Creative Solution
23:04 • 4min
10
The Importance of Alignment in Language Models
26:43 • 4min
11
The Intrinsic Difficulties of Catastrophic Risk From AI
30:54 • 4min
12
The Importance of Inner Alignment in AI
34:55 • 2min
13
The Importance of Goals in Evolution
37:06 • 4min
14
How to Maximize Your Debate
40:39 • 5min
15
The Power of Smiley Faces
45:38 • 4min
16
The Role of Alignment in AI Safety
49:57 • 5min
17
The Importance of Conciseness in AI Safety
55:20 • 1min
18
The Pros and Cons of the Cheese Vector
56:38 • 5min
19
The Importance of Inference in Deep Neural Networks
01:01:59 • 3min
20
The Importance of Socializing With Government
01:04:47 • 5min
21
The Importance of Iterating on Messaging
01:09:54 • 4min
22
The Future of Auto GPT
01:13:47 • 5min
23
The Open Source Limitation of Chaos GPT
01:18:27 • 2min
24
Quantum Mechanics Made Me Do It
01:20:24 • 4min
25
The Weird Things About Quantum Mechanics
01:23:59 • 7min