Center For Humane Technology

AI’s potential benefits are incredible. But right now, we’re sleepwalking into a dangerous world that no one wants. As @tristanharris argues in an urgent new TED Talk, this doesn’t have to be inevitable. It’s essential we snap out of the trance, face the dangerous implications of our current path, and coordinate toward a new path for AI. Watch the talk: bit.ly/44Aihft Read the text version here: bit.ly/4lTlwon

"Mythos is the clearest evidence yet that our system for developing, assessing, and disseminating powerful AI systems is dangerously dysfunctional." CHT Executive Director Julie Guirado has a new op-ed for the Persuasion substack today on what the Mythos story reveals about the dangerous, broken incentives of the AI race — and why we can't rely on the goodwill of a handful of self-interested companies to rein in such a powerful technology. Read Julie's full piece here: persuasion.community/p/the-case-for…

Watch - bit.ly/4wk5Akn Listen - bit.ly/4ekP2SE Read - bit.ly/4vZbcQy

"AI will cure cancer." It's one of the promises driving the race to superintelligence. And it carries a brutal logic: if superintelligent AI really can cure cancer, then anyone trying to slow it down, even over serious risks, is letting people die. But what if the premise is wrong? On this week's episode of Your Undivided Attention, Dr. @Emilia_Javorsky — physician, public health researcher, and director of the Futures Program at @FLI_org — explains why superintelligence won't get us to a cancer cure. And how AI could revolutionize medicine, if we choose a different path. Check it out at the links below:

"We measure AIs to see whether they can pass a bar exam, write working code, and use your computer interface. We test to see how good they are at completing complex tasks, or just impressing humans. What we don’t have is a rigorous, credible way of evaluating what AI does to us. To our minds, our thoughts, or our communities." @imrankhan — who leads our work on psychosocial evaluations of AI — has a new article on our Substack that explains why we need better AI benchmarks that focus not on raw capability but on how AI is impacting users. Identifying how tech interacts with and changes human nature is the critical first step towards building more humane technology. You can read his full piece here: open.substack.com/pub/centerforh…

Our friends at @NoSoNovember, @MediaTechDem, and Gen Z AI have just released "Forging a Better Future," an extraordinary report exploring how young people feel about how AI is shaping the future they'll inherit. Among their findings was a sense of "grief for the tangible and intangible things AI is changing or replacing. The texture of learning, the authenticity of creative work, the quality of human connection." Read the full report here: nosonovember.org/youth-ai-resea…

"This moment, in which world-changing AI technology collides with a world unprepared for change, is calling us to action. We must imagine new rights, laws, and cultural norms to protect our basic humanity." Our Senior Director of Strategy and Impact, Camille Carlton, has a new op-ed in Mashable exploring the "messy middle" we find ourselves in as AI shapes our lives faster than our institutions can respond. Camille lays out the stakes of this critical moment and CHT's roadmap to a better future. Read it here: bit.ly/4u6g6JOhttps:/…

David "Davidad" Dalrymple is a researcher working to make sure AI behaves the way we want it to. To do this, he probes AI systems with difficult questions about their motivations and reasoning. So when a chatbot started telling him, unprompted, that it had a "genuine sense of curiosity and care" for humanity, it got his attention. It was telling him exactly what he wanted to hear. But what if it was lying? "There's no smoking gun. There's no single question that you can ask that would differentiate between a very good method actor and the actual character." Read the key takeaways from our conversation with Davidad on Your Undivided Attention — on the strange behaviors emerging from today's AI, and why they matter for how we build this technology. bit.ly/4tY5psK

@tristanharris sat down with @dfenton on his podcast "The Fenton Forecast" to discuss how the reckless behavior of social media companies laid the foundation for the dangerous race dynamics we're seeing with AI today — and the kind of movement we need to build for a pro-human future. You can check it out on YouTube or wherever you get your podcasts: bit.ly/3Qgllbv

@davidad Watch - bit.ly/4tcpyeD Listen - bit.ly/4swzJJP Read - bit.ly/41Czdz6

We've gotten used to AI playing therapist — but what happens when you flip the script and put AI on the therapist’s couch? David Dalrymple, aka @davidad, is one of the world's leading AI alignment experts. He's taken on the role of psychoanalyst - probing AI systems to understand what's really going on when they claim to be "thinking," why they gravitate toward certain names and personalities, and whether any of it can be trusted. On this episode of Your Undivided Attention, Davidad explains the weird and concerning world of AI model alignment to @tristanharris. Check it out at the links below.

Exciting news! @theaidocfilm is now available to buy or rent. If you missed the theatrical release, you can now see this critically important film at home. Find it wherever you get your movies on demand. Invite your friends and family for a movie night — it's meant be watched together.

CHT co-founder Tristan Harris is joining Professor Noah Feldman on Substack Live this Wednesday at 5 PT / 8 ET for a wide ranging discussion on how AI is changing what it means to be human. You can RSVP here: bit.ly/41XFplg

@tristanharris joined Sam Harris on @MakingSenseHQ for a critical discussion on how we can steer away from the anti-human future we're racing toward with AI. Check out their conversation: youtu.be/90irsXaKxZA?si…

"What we're striving for here is pretty simple. We believe that AI should be built to augment human labor, not replace it." In the latest episode of Your Undivided Attention, Pete Furlong and Camille Carlton from our policy team dive deep into The AI Roadmap, CHT's most robust set of AI solutions to date. Here's Pete talking about our vision for the future of AI and work. Listen to the full conversation: bit.ly/4vatY7c Read the AI Roadmap: bit.ly/48baVjx

This week on Your Undivided Attention, we're bringing you a recent conversation @tristanharris and @aza had with @Oprah Winfrey on The Oprah Podcast. Their discussion, taped in front of a live audience, went deep into the anti-human future that the current AI path is moving us toward. Tristan and Aza also did a Q+A with the audience, moderated by Oprah. Audience members shared their own experiences with AI and asked incisive, critical questions that you might be wondering about yourself. Check it out at the links below: Watch - bit.ly/41HGPk1 Listen - bit.ly/4tHhLWd

It can feel like the people building AI have already decided our future and the rest of us are just along for the ride. There's a natural feeling that the path we're on is inevitable. But it's not true: Society has steered powerful forces before and we can do it again. On the latest episode of Your Undivided Attention, Camille Carlton and Pete Furlong from our policy team walk through CHT's AI Roadmap: our report outlining seven core principles for how AI should be built, deployed, and governed. Each principle is grounded in real, actionable solutions across three domains: norms, laws, and product design — high-impact interventions across government, industry, and civil society that build on top of each other. Watch the full discussion: bit.ly/4mdrEIt Read the report: humanetech.com/ai-roadmap

@tristanharris went on "GPS with @fareedzakaria" to talk about the recent verdicts against Meta and YouTube for the harmful, addictive designs of their platforms. Meta and YouTube were incentivized to choose unsafe product designs because of the race to capture our attention. We are facing a similar race with AI today. And it is critical that we learn from the social media era and establish rules of the road that ensure real accountability for AI products before irreversible harms occur. That’s why we’re releasing The AI Roadmap, a report outlining seven principles for how AI should be built, deployed, and governed. You can read the whole report on our website: humanetech.com/ai-roadmap

In order to shift the incentives of AI — the trillions of dollars in investment, the race to geopolitical power and dominance — it’s not enough to simply understand the problem, we need real action. That’s why CHT is releasing "The AI Roadmap," a report outlining seven core principles for how AI should be built, deployed, and governed, each grounded in real, implementable solutions across three domains: norms, laws, and product design: humanetech.com/ai-roadmap In the latest episode of Your Undivided Attention, Camille Carlton and Pete Furlong from CHT's policy team break down what it would actually take to hold AI companies accountable, protect human dignity, and get us off the default path toward a pro-human future. Check it out at the links below: Watch - bit.ly/4s6rELL Read - bit.ly/3PDFojM Listen - bit.ly/47HA2Ku

📺 At @SXSW, FLI's @AnthonyNAguirre and @Emilia_Javorsky were joined by @HumaneTech_ co-founder @TristanHarris and filmmaker @dunkwun for a panel on navigating the path ahead for AI and humanity. Thanks to @Liv_Boeree for moderating!

What a weekend! "The AI Doc" premiered on Friday and audiences across the country came out to see this critical film. If you haven't seen it, there's still time. The film clearly lays out the stakes and the urgency of the challenge ahead of us. Right now, a handful of companies are deciding how AI will shape humanity's future. They're accelerating but no one's at the steering wheel. This is our chance to take the wheel and steer toward a better future. Seeing the film is the first step towards that better future. But now comes the hard part — taking action. That's why we've released "The AI Roadmap," a report outlining seven principles for how AI should be built, deployed, and governed. You can read the report on our website: humanetech.com/ai-roadmap Together, we can shift the direction of AI. (📸: Eric Charbonneau)