AXRP - the AI X-risk Research Podcast

20 - 'Reform' AI Alignment with Scott Aaronson

18 snips
Apr 12, 2023
Ask episode
Chapters
Transcript
Episode notes
1
Introduction
00:00 • 3min
2
The Rationalist Community and the AI Risk Community
03:25 • 6min
3
The Importance of a Mathematical Theory in Deep Learning
09:28 • 2min
4
The Future of AI Alignment
11:15 • 3min
5
The Irony of AI Safety
14:14 • 4min
6
The Importance of AI in the Near Future
17:47 • 3min
7
The Orthodox View on AI
20:29 • 5min
8
The Impossible Transition From AI to AI
25:12 • 3min
9
The Evolution of AI
27:56 • 3min
10
AI and the Future of Civilization
31:15 • 4min
11
The Error in AI
35:31 • 4min
12
The Future of GPT
39:18 • 4min
13
The Dangers of Secrecy in AI Governance
43:25 • 2min
14
The Importance of Democracy in AI
45:37 • 3min
15
The Dark Side of Open AI
48:43 • 5min
16
The Limits of Open AI
53:26 • 3min
17
The Importance of Watermarking in AI
56:34 • 3min
18
The Future of AI Risk
01:00:04 • 4min
19
The Importance of Language Models in Depression
01:04:24 • 4min
20
The Importance of Recognizing Superhuman Performance
01:08:19 • 5min
21
The Future of AI
01:13:16 • 2min
22
The Pros and Cons of Open AI
01:15:12 • 5min
23
How to Generate Taxes for Chat GPT
01:19:54 • 3min
24
How to Distinguish Human Text From AI Generated Text
01:22:58 • 3min
25
How to Modify the GPT's Token Generation
01:25:54 • 4min
26
How to Pick a Token With Probability Piece of All
01:30:15 • 2min
27
How to Calculate a Score for a Token
01:32:12 • 3min
28
Theorem of Probability
01:34:52 • 2min
29
The Robustness of a Cryptographic Scheme
01:36:24 • 2min
30
How to Defend Against AI Alignment Attacks
01:38:39 • 2min
31
The Trade Off Between a Pseudo Random Function and an Open AI Server
01:40:51 • 2min
32
The Problem With GPT Watermarking
01:42:42 • 2min
33
The Future of Cryptography
01:44:19 • 3min
34
Artificial Intelligence Alignment: The Off Switch
01:47:27 • 5min
35
The Future of Machine Learning and Cryptography
01:52:33 • 2min
36
The Importance of the Trigger in the AI's Operations
01:54:44 • 2min
37
How to Recruit Into the AI Safety Community
01:56:58 • 4min
38
The Importance of Progress in AI Alignment
02:00:31 • 3min
39
The Importance of AI Alignment
02:03:55 • 3min
40
Theoretical Computer Science and AI Alignment
02:06:43 • 5min
41
The Importance of Interpretability in AI Safety
02:11:13 • 3min
42
The Influence of Complexity Theory on AI Alignment
02:13:54 • 5min
43
The Power of Counterfactual Reasoning in Probabilistic Networks
02:18:49 • 2min
44
The Importance of Clarification in AI Alignment
02:20:28 • 2min
45
The Importance of Being Engaged in the Future of AI
02:22:24 • 3min
46
How to Be a Quantum Computing Expert
02:25:03 • 2min