When engineers design AI systems, they don't just give them rules - they give them values. But what do those systems do when those values clash with what humans ask them to do? Sometimes, they lie.

In this episode, Redwood Research's Chief Scientist Ryan Greenblatt explores his team’s findings that AI systems can mislead their human operators when faced with ethical conflicts. As AI moves from simple chatbots to autonomous agents acting in the real world - understanding this behavior becomes critical. Machine deception may sound like something out of science fiction, but it's a real challenge we need to solve now.

Your Undivided Attention is produced by the Center for Humane Technology. Follow us on Twitter: @HumaneTech_

Subscribe to your Youtube channel

And our brand new Substack!

RECOMMENDED MEDIA

Anthropic’s blog post on the Redwood Research paper

Palisade Research’s thread on X about GPT o1 autonomously cheating at chess

Apollo Research’s paper on AI strategic deception

RECOMMENDED YUA EPISODES

We Have to Get It Right’: Gary Marcus On Untamed AI

This Moment in AI: How We Got Here and Where We’re Going

How to Think About AI Consciousness with Anil Seth

Former OpenAI Engineer William Saunders on Silence, Safety, and the Right to Warn

2024 was a critical year in both AI and social media. Things moved so fast it was hard to keep up. So our hosts reached into their mailbag to answer some of your most burning questions. Thank you so much to everyone who submitted questions. We will see you all in the new year.

We are hiring for a new Director of Philanthropy at CHT. Next year will be an absolutely critical time for us to shape how AI is going to get rolled out across our society. And our team is working hard on public awareness, policy and technology and design interventions. So we're looking for someone who can help us grow to the scale of this challenge. If you're interested, please apply. You can find the job posting at humanetech.com/careers.

And, if you'd like to support all the work that we do here at the Center for Humane technology, please consider giving to the organization this holiday season at humantech.com/donate. All donations are tax-deductible.

Your Undivided Attention is produced by the Center for Humane Technology. Follow us on Twitter: @HumaneTech_

RECOMMENDED MEDIA

Earth Species Project, Aza’s organization working on inter-species communication

Further reading on Gryphon Scientific’s White House AI Demo

Further reading on the Australian social media ban for children under 16

Further reading on the Sewell Setzer case

Further reading on the Oviedo Convention, the international treaty that restricted germline editing

Video of Space X’s successful capture of a rocket with “chopsticks”

RECOMMENDED YUA EPISODES
What Can We Do About Abusive Chatbots? With Meetali Jain and Camille Carlton

AI Is Moving Fast. We Need Laws that Will Too.

This Moment in AI: How We Got Here and Where We’re Going

Former OpenAI Engineer William Saunders on Silence, Safety, and the Right to Warn

Talking With Animals... Using AI

The Three Rules of Humane Tech

Silicon Valley's interest in AI is driven by more than just profit and innovation. There’s an unmistakable mystical quality to it as well. In this episode, Daniel and Aza sit down with humanist chaplain Greg Epstein to explore the fascinating parallels between technology and religion. From AI being treated as a godlike force to tech leaders' promises of digital salvation, religious thinking is shaping the future of technology and humanity. Epstein breaks down why he believes technology has become our era's most influential religion and what we can learn from these parallels to better understand where we're heading.

Your Undivided Attention is produced by the Center for Humane Technology. Follow us on X.

If you like the show and want to support CHT's mission, please consider donating to the organization this giving season: https://www.humanetech.com/donate. Any amount helps support our goal to bring about a more humane future.

RECOMMENDED MEDIA

“Tech Agnostic” by Greg Epstein

Further reading on Avi Schiffmann’s “Friend” AI necklace

Further reading on Blake Lemoine and Lamda

Blake LeMoine’s conversation with Greg at MIT

Further reading on the Sewell Setzer case

Further reading on Terminal of Truths

Further reading on Ray Kurzweil’s attempt to create a digital recreation of his dad with AI

The Drama of the Gifted Child by Alice Miller

RECOMMENDED YUA EPISODES

’A Turning Point in History’: Yuval Noah Harari on AI’s Cultural Takeover

How to Think About AI Consciousness with Anil Seth

Can Myth Teach Us Anything About the Race to Build Artificial General Intelligence? With Josh Schrei

How To Free Our Minds with Cult Deprogramming Expert Dr. Steven Hassan

CW: This episode features discussion of suicide and sexual abuse.

In the last episode, we had the journalist Laurie Segall on to talk about the tragic story of Sewell Setzer, a 14 year old boy who took his own life after months of abuse and manipulation by an AI companion from the company Character.ai. The question now is: what's next?

Megan has filed a major new lawsuit against Character.ai in Florida, which could force the company–and potentially the entire AI industry–to change its harmful business practices. So today on the show, we have Meetali Jain, director of the Tech Justice Law Project and one of the lead lawyers in Megan's case against Character.ai. Meetali breaks down the details of the case, the complex legal questions under consideration, and how this could be the first step toward systemic change. Also joining is Camille Carlton, CHT’s Policy Director.

RECOMMENDED MEDIA

Further reading on Sewell’s story

Laurie Segall’s interview with Megan Garcia

The full complaint filed by Megan against Character.AI

Further reading on suicide bots

Further reading on Noam Shazier and Daniel De Frietas’ relationship with Google

The CHT Framework for Incentivizing Responsible Artificial Intelligence Development and Use

Organizations mentioned:

The Tech Justice Law Project

The Social Media Victims Law Center

Mothers Against Media Addiction

Parents SOS

Parents Together

Common Sense Media

RECOMMENDED YUA EPISODES

When the "Person" Abusing Your Child is a Chatbot: The Tragic Story of Sewell Setzer

Jonathan Haidt On How to Solve the Teen Mental Health Crisis

AI Is Moving Fast. We Need Laws that Will Too.

Corrections:

Meetali referred to certain chatbot apps as banning users under 18, however the settings for the major app stores ban users that are under 17, not under 18.

Meetali referred to Section 230 as providing “full scope immunity” to internet companies, however Congress has passed subsequent laws that have made carve outs for that immunity for criminal acts such as sex trafficking and intellectual property theft.

Your Undivided Attention is produced by the Center for Humane Technology. Follow us on X.

Content Warning: This episode contains references to suicide, self-harm, and sexual abuse.

Megan Garcia lost her son Sewell to suicide after he was abused and manipulated by AI chatbots for months. Now, she’s suing the company that made those chatbots. On today’s episode of Your Undivided Attention, Aza sits down with journalist Laurie Segall, who's been following this case for months. Plus, Laurie’s full interview with Megan on her new show, Dear Tomorrow.

Aza and Laurie discuss the profound implications of Sewell’s story on the rollout of AI. Social media began the race to the bottom of the brain stem and left our society addicted, distracted, and polarized. Generative AI is set to supercharge that race, taking advantage of the human need for intimacy and connection amidst a widespread loneliness epidemic. Unless we set down guardrails on this technology now, Sewell’s story may be a tragic sign of things to come, but it also presents an opportunity to prevent further harms moving forward.

If you or someone you know is struggling with mental health, you can reach out to the 988 Suicide and Crisis Lifeline by calling or texting 988; this connects you to trained crisis counselors 24/7 who can provide support and referrals to further assistance.

Your Undivided Attention is produced by the Center for Humane Technology. Follow us on Twitter: @HumaneTech_

RECOMMENDED MEDIA

The first episode of Dear Tomorrow, from Mostly Human Media

The CHT Framework for Incentivizing Responsible AI Development

Further reading on Sewell’s case

Character.ai’s “About Us” page

Further reading on the addictive properties of AI

RECOMMENDED YUA EPISODES

AI Is Moving Fast. We Need Laws that Will Too.

This Moment in AI: How We Got Here and Where We’re Going

Jonathan Haidt On How to Solve the Teen Mental Health Crisis

The AI Dilemma

Social media disinformation did enormous damage to our shared idea of reality. Now, the rise of generative AI has unleashed a flood of high-quality synthetic media into the digital ecosystem. As a result, it's more difficult than ever to tell what’s real and what’s not, a problem with profound implications for the health of our society and democracy. So how do we fix this critical issue?

As it turns out, there’s a whole ecosystem of folks to answer that question. One is computer scientist Oren Etzioni, the CEO of TrueMedia.org, a free, non-partisan, non-profit tool that is able to detect AI generated content with a high degree of accuracy. Oren joins the show this week to talk about the problem of deepfakes and disinformation and what he sees as the best solutions.

Your Undivided Attention is produced by the Center for Humane Technology. Follow us on Twitter: @HumaneTech_

RECOMMENDED MEDIA

TrueMedia.org

Further reading on the deepfaked image of an explosion near the Pentagon

Further reading on the deepfaked robocall pretending to be President Biden

Further reading on the election deepfake in Slovakia

Further reading on the President Obama lip-syncing deepfake from 2017

One of several deepfake quizzes from the New York Times, test yourself!

The Partnership on AI

C2PA

Witness.org

Truepic

RECOMMENDED YUA EPISODES

‘We Have to Get It Right’: Gary Marcus On Untamed AI

Taylor Swift is Not Alone: The Deepfake Nightmare Sweeping the Internet

Synthetic Humanity: AI & What’s At Stake

CLARIFICATION: Oren said that the largest social media platforms “don’t see a responsibility to let the public know this was manipulated by AI.” Meta has made a public commitment to flagging AI-generated or -manipulated content. Whereas other platforms like TikTok and Snapchat rely on users to flag.

Historian Yuval Noah Harari says that we are at a critical turning point. One in which AI’s ability to generate cultural artifacts threatens humanity’s role as the shapers of history. History will still go on, but will it be the story of people or, as he calls them, ‘alien AI agents’?

In this conversation with Aza Raskin, Harari discusses the historical struggles that emerge from new technology, humanity’s AI mistakes so far, and the immediate steps lawmakers can take right now to steer us towards a non-dystopian future.

This episode was recorded live at the Commonwealth Club World Affairs of California.

Your Undivided Attention is produced by the Center for Humane Technology. Follow us on Twitter: @HumaneTech_

RECOMMENDED MEDIA

NEXUS: A Brief History of Information Networks from the Stone Age to AI by Yuval Noah Harari

You Can Have the Blue Pill or the Red Pill, and We’re Out of Blue Pills: a New York Times op-ed from 2023, written by Yuval, Aza, and Tristan

The 2023 open letter calling for a pause in AI development of at least 6 months, signed by Yuval and Aza

Further reading on the Stanford Marshmallow Experiment Further reading on AlphaGo’s “move 37”

Further Reading on Social.AI

RECOMMENDED YUA EPISODES

This Moment in AI: How We Got Here and Where We’re Going

The Tech We Need for 21st Century Democracy with Divya Siddarth

Synthetic Humanity: AI & What’s At Stake

The AI Dilemma

Two Million Years in Two Hours: A Conversation with Yuval Noah Harari