Reflections on Intelligence

by Magnus Vinding

A lot of people are talking about “superintelligent AI” these days. But what are they talking about? Indeed, what is “intelligence” in the first place? This question is generally left unanswered, even unasked, in discussions about the perils and promises of artificial intelligence, which tends to make these discussions more confusing than enlightening. More clarity and skepticism about this term “intelligence” is desperately needed. Hence this book. 'Reflections on Intelligence' aims to look deeper into the concept of “intelligence”, and to question common assumptions about the phenomenon, or rather phenomena, that we refer to with the word “intelligence”. Based on such an examination, it proceeds to criticize some of the views and ideas commonly discussed in relation to the future of artificial intelligence, and to draw some general conclusions about the future of "intelligence".

76 pages, Paperback

Published May 2, 2017

About the author

Magnus Vinding is the author of Speciesism: Why It Is Wrong and the Implications of Rejecting It (2015), Reflections on Intelligence (2016), You Are Them (2017), Effective Altruism: How Can We Best Help Others? (2018), Suffering-Focused Ethics: Defense and Implications (2020), Reasoned Politics (2022), Essays on Suffering-Focused Ethics (2022), and Essays on UFOs and Related Conjectures (2024).

His next book will be Compassionate Purpose: Personal Inspiration for a Better World.

Reviews

[Teo · Rating 4 out of 5]

2018.10.24–2018.10.24

Contents

Vinding M (2016) Reflections on Intelligence

Introduction

1. What is Intelligence?
• Origin and Common Use of the Term
• A Clear Definition of a Relevant Ability

2. Goal Achieving – Its Distributed and Many-Faceted Nature
• The Relationship Between Individual and Collective Goal Achieving Ability

3. When Machines Improve Machines
• “Intelligence Though!” – A Bad Argument
• The Problems with Magic Sauce Theory

4. Consciousness – Orthogonal or Crucial?

5. A Brief Note on Goals

6. The Unpredictability of the Future of “Intelligence”

Conclusion

Acknowledgments

Bibliography

[Daniel Hageman · Rating 4 out of 5]

This one definitely provides reason to recalibrate one's risk assessment of the issues surrounding AI and AI alignment, largely by highlighting just how nebulous the concept of 'intelligence' tends to be in many of the more typical claims in this area. Will definitely need to come back to this one after looking again at Bostrom's work and others, and see to what extent I am indeed prioritizing the issue with the level of importance it deserves (or doesn't). Or at the very least, ensuring I have a more refined idea of what 'the problem' entails.