The Lawfare Podcast

Cybersecurity and AI

Apr 3, 2023

01:04:37

Creator website

Episode guests

Nicole Perlroth

Alex Stamos

Dave Willner

Highlights

AI Chapters

Episode notes

Cybersecurity of AI: A Multifaceted Challenge

01:37

AI and Cybersecurity Risks

02:19

Content Moderation in LLMs

01:40

Misconceptions about AI sentience

02:53

Introduction

3min

The Cybersecurity of Artificial Intelligence

2min

The PLA: A Space Full of Ideas

2min

The Risks of Model Weights in AI

2min

The Importance of Model Weights in AI

2min

How to Stop Paying Attention to Cyber Security and Focus on Vanderpump Rules

2min

The Threat of Artificial Intelligence

2min

The Theft of AI

3min

The Threat to the Algorithms

2min

10.

How to Hack Chat GPT's Content Moderation

2min

11.

How to Use Twitter's Model to Moderate Content

2min

12.

How to Use Microsoft's Content Moderation to Improve Your Code

2min

13.

How Worried Are You About the Conduct and Controllability of the Algorithms Once They're Loosed on the World?

2min

14.

The Dangers of AI and Machine Learning

3min

15.

How to Do Multi Step Tasks

3min

16.

The Importance of Autonomy in AI

2min

17.

The Future of AI

2min

18.

The Threat of Human Misuse of Chat GPT

3min

19.

How Chatbots Can Help Influence Politics

3min

20.

How to Prevent Disinformation From Become a Vehicle

2min

21.

The Problem With Disinformation

4min

22.

The Power of Consciousness

2min

23.

Chat TPT and Other LLMs Don't Have a Soul

2min

24.

The Importance of Models in AI Research

3min

25.

The Problem With Regulating OpenAI

2min

26.

The Importance of External Auditing and Red Teaming

3min

27.

How to Defy Nicole's Fear of Skynet

2min

28.

The Future of Malware Prevention

3min

29.

Cucumbers: A Variety of Uses

3min

Hosted by the William and Flora Hewlett Foundation Cyber Initiative and Aspen Digital, Verify 2023 brings together journalists and cyber and tech policy experts to discuss critical issues in cybersecurity. For this live recording of the Lawfare Podcast, Benjamin Wittes sat down at Verify 2023 with Alex Stamos of the Stanford Internet Observatory; Nicole Perlroth, formerly of the New York Times and the author of a recent book on zero days; and Dave Willner, the Head of Trust & Safety at OpenAI, the company that produces ChatGPT. They talked about cybersecurity and AI, the threats to AI algorithms, the threats from AI algorithms, and the threats from humans misusing large language models.

Support this show http://supporter.acast.com/lawfare.

Hosted on Acast. See acast.com/privacy for more information.