The Gradient: Perspectives on AI

Melanie Mitchell: Abstraction and Analogy in AI

Dec 15, 2022

54:47

Creator website

Episode guests

Melanie Mitchell

Highlights

AI Chapters

Episode notes

Analogy in Everyday Life

02:24

The Evolution of Statistical Machine Learning

01:52

The First Step Fallacy in AI

02:22

The Limits of Statistical Learning in Image Recognition

01:57

Introduction

2min

How Did You Get Into AI?

2min

How Does Not Having a Computer Science Background Influence Your AI Research?

2min

Is There a Difference Between Symbolic AI and Machine Learning?

2min

Is Scaling the Next Big Step in AI Progress?

2min

How Do We Evaluate Language Models?

3min

The Chinese Room and the Zombies

2min

Why AI Is Harder Than We Think?

5min

Is There a Difference Between Human and Computer?

3min

10.

Do You Want to Replicate Human-Like Intelligence?

2min

11.

What Is Intelligence?

2min

12.

Is There a Degree of Intelligence?

2min

13.

Is It Even Useful to Slap That Label on Intelligence?

2min

14.

Is There a Need for Scaling in Artificial Intelligence?

2min

15.

Is There a Causality in the Text to Image Systems?

2min

16.

Analogy and Machine Learning

3min

17.

A Concept Is Something Much Richer Than a Category

2min

18.

Is There a Place That Bottoms Out?

3min

19.

Is There a Bottoming Out in the Physical?

2min

20.

Copycat Architecture - What Is It?

3min

21.

Neural Networks and Symbolic AI

2min

22.

What Advice Would You Give to Someone Starting Out in AI?

4min

Have suggestions for future podcast guests (or other feedback)? Let us know here!

In episode 53 of The Gradient Podcast, Daniel Bashir speaks to Professor Melanie Mitchell.

Professor Mitchell is the Davis Professor at the Santa Fe Institute. Her research focuses on conceptual abstraction, analogy-making, and visual recognition in AI systems. She is the author or editor of six books and her work spans the fields of AI, cognitive science, and complex systems. Her latest book is Artificial Intelligence: A Guide for Thinking Humans.

Subscribe to The Gradient Podcast: Apple Podcasts | Spotify | Pocket Casts | RSSFollow The Gradient on Twitter

Outline:

* (00:00) Intro

* (02:20) Melanie’s intro to AI

* (04:35) Melanie’s intellectual influences, AI debates over time

* (10:50) We don’t have the right metrics for empirical study in AI

* (15:00) Why AI is Harder than we Think: the four fallacies

* (20:50) Difficulties in understanding what’s difficult for machines vs humans

* (23:30) Roles for humanlike and non-humanlike intelligence

* (27:25) Whether “intelligence” is a useful word

* (31:55) Melanie’s thoughts on modern deep learning advances, brittleness

* (35:35) Abstraction, Analogies, and their role in AI