The Gradient: Perspectives on AI

Yejin Choi: Teaching Machines Common Sense and Morality

May 19, 2022
Ask episode
Chapters
Transcript
Episode notes
1
Introduction
00:00 • 2min
2
Are We Going to Have Another Winter?
02:20 • 2min
3
What's the Difference Between Deep Learning and NLP?
03:59 • 1min
4
Theoretical Models and Theorems
05:29 • 2min
5
Natural Language Processing Is Chasing After the Wrong Goal?
07:29 • 4min
6
What's the Bottom Up Approach to Learning From the Data?
11:17 • 3min
7
Common Sensitivity in Machine Learning
14:14 • 3min
8
Is There a Machine's Ability to Pick Up the World?
16:59 • 3min
9
The Concept of Abduction or Obductive Reasoning
20:13 • 3min
10
Common Sense Reasoning in Multiple Choice Questions
23:43 • 3min
11
Multiple Choice Questions Aren't Superindicative of Common Sense Intelligence
26:52 • 3min
12
Is There a Difference Between Generative and Discriminati Valuation?
29:29 • 3min
13
How to Improve Predict Middle Sentence by Reading Left and Right Sentences
32:13 • 4min
14
Dolorian and Thes Inferance Base Common Sense
36:06 • 4min
15
Taking a Pre Trained Language Model and Filtering Out the Bad Information?
40:07 • 3min
16
Common Sensible Reasoning in Language Models
42:47 • 3min
17
The Importance of Common Sensible Reasoning
46:03 • 3min
18
The Ethical Implications of Common Sense Intelligence
48:36 • 4min
19
How Do You Measure the Performance of a Language Model?
52:36 • 4min
20
How Do Humans Evaluate and Make Moral Judgments Using Descriptive Morality?
56:22 • 4min
21
Is It a Can of Worms?
01:00:12 • 3min
22
Are Tatai Systems a Moral Authority or Sources of Advice on Human Ethics?
01:03:08 • 4min
23
I Think the Policy Factors in the Cultural Norms Are Really Important
01:07:11 • 4min
24
What's Next for Computer Science?
01:10:46 • 4min
25
Gradiand Podcast - Part 2
01:14:18 • 2min