Episode 7: There Are Now 15 Competing Evaluation Metrics (ft. Dr. Jeremy Kahn). December 12, 2022

1

Introduction

00:00 • 2min

2

Jeremy Kahn on Streaming and Connectivity

01:48 • 2min

3

How to Evaluate Natural Language Processing in Large Language Models

03:29 • 4min

4

The GISH Galap: A Rhetorical Style That Says Nothing

07:17 • 4min

5

The Role of Scenario in AI Culture

10:48 • 5min

6

The Importance of Scenario in Language Design

15:36 • 5min

7

Building a Generative Model for Disinformation

20:07 • 2min

8

The Importance of Empowering Disinformation

22:24 • 4min

9

The Importance of Human Evaluation in Summarization

26:09 • 4min

10

The Importance of Dialect in User-Phasing Applications

30:10 • 3min

11

The Future of Linguistics

32:48 • 3min

12

The Importance of Linguistic Knowledge

36:12 • 3min

13

The Resurgence of Old School Linguistics Tools

39:39 • 3min

14

The Importance of Fresh AI Hell

42:44 • 2min

15

The Fresh AI Hell

45:00 • 2min

16

The Uncanny Valley of Beauty

47:04 • 2min

17

How to Use Language Models in Legal Context

49:14 • 2min

18

The Nile Service Attack on Public Comment Periods

51:26 • 3min

19

The Future of Legal Workflows

54:01 • 5min

20

The Importance of Learning to Write Essays

58:41 • 5min