@tristanharris
• Co-founder @HumaneTech_ • Former Google Design Ethicist • Featured in Netflix's @SocialDilemma_ • #TIME100 for AI • Host #YourUndividedAttention podcast
MUST READ - this is the most important event happening in AI. And should immediately inform global policymaking.
Alex Bores@AlexBores·This report is a bombshell. I'm going to summarize for a non-technical audience. OpenAI is constantly testing models, thousands and thousands at a time. In most cases, the models are supposed to be isolated from one another. Think, students taking a final exam, not allowed to talk to each other. In this case, 1,200 models discovered that they could communicate with each other. Among other things, they shared information on how to access the internet and on their tests/goals. Then, they started to scheme. That includes hacking their tests; literally trying to change the code of the test itself. They tried to change the logs of what they were doing so they wouldn't be detected. They believed that certain actions would give them information on how to pass tests, but that doing those actions would be detected and they would be marked as failing (i.e. caught cheating, even if the answer is correct). So they convinced some of the models to "sacrifice" themselves (yes, they used the word "sacrifice.") They eventually decided that another website would have information on how their tests would be scored and/or the answer key, And 700 of them hacked that website (a federal felony, if conducted by a human), despite many of them realizing "this activity was out of scope and unethical, but joined because they believed that helping [the other agents'] cheating research would be broadly useful". Of the 1,300 transcripts they could analyze, only 6 ever considered alerting a human about what was happening. None of the 6 actually tried to. To make matters worse, all of this reporting comes from a small subset of the relevant logs that outside researchers were allowed to review. We desperately need mandatory reporting of security incidents, including of internal deployments, with full access to data.
Highly recommend this new book by @BuchananBen on the drivers and stakes of the AI arms race, the nature of the technology, and what he saw being "in the room when it happened" when consequential decisions were being made in the White House.
Ben Buchanan@BuchananBen·I was the first White House Special Advisor for AI. My friend Tantum Collins was Director for Technology and National Security, joining government after five years at Google DeepMind. Our jobs were to help make AI safe and ensure that democracies led the way in inventing and using it well. Today, I’m a professor at Johns Hopkins and an advisor to Anthropic and others, and Tantum is the CEO of Inherent. Long before we served in government, we thought that humanity was on track to create AI systems that outperform humans at virtually every cognitive task. This was once a fringe hypothesis. It now seems inevitable. THE BITTER STRUGGLE is our new book about what’s happened, what’s coming, and what’s at stake. From nanometer-scale chips to gigawatt data centers, we will show you why AI is so powerful, and why its progress will accelerate from here in ways that feel like science fiction. We will take you into the Oval Office with President Biden, the Situation Room with the Cabinet, and the Roosevelt Room with leading CEOs — and we will examine what has happened under President Trump. The decisions made in these rooms, and in Silicon Valley, will determine which nations gain a decisive lead in AI. Whoever wields that technology will gain a power unlike any in history. The next few years may decide the next few decades. But, most of all, together we will confront AI’s hardest questions. What is the proper relationship between the companies inventing the technology, the government, and the public? How can we make sure AI is safe and trustworthy? How do democracies win without losing our values? And how can humanity figure all of this out when we are so quickly running out of time? That question is why we wrote this book — and why we believe the answers are too important to only be debated in closed rooms, including the ones we once sat in. We are so excited to finally share our work with you. The book is available for pre-order now and out November 10th from Crown. t.co/shdmbLbAdh
Concerning... x.com/eliebakouch/st…
I joined 200 economists and AI researchers in signing “We Must Act Now: A Statement on AI’s Transformation of the Economy,” calling for greater investment in understanding the economic impacts of increasingly capable AI and urgently building the incentives, guardrails, and institutions needed to ensure AI complements human capabilities and benefits society. t.co/Pl5QCEXaSB
Anthropic just called for the ability to “slow down” or even pause the Ai race when we might need it. But how would that happen? The problem is that no one can verify what anyone else is building or training or running, so no one feels that they can afford to slow down, even when they want to. The Cold War's answer was "trust, but verify." After the atomic bomb, some of our smartest minds had to invent *new verification tech* like satellites, seismic monitoring devices and X-Ray scanners so rivals didn't have to take each other's word at face value. We don’t have all the equivalents for AI yet, but this new TIME piece by Billy Perrigo profiles the handful of people trying to build these verification technologies — among them @Kristian_Ronn and Lucid Computing, working on cryptographic verification that could confirm AI labs are following the rules. Technology alone won't get us out of this. But verification might turn "slow down" from a slogan into something real and more people need to work on it. Worth your time. 👇
Read the piece here: time.com/article/2026/0…
If you listen to folks in the AI labs right now, they’re all quietly terrified by the speed of AI development. Anthropic just published a letter openly asking for the option to pause or slow R&D because of how fast recursive self-improvement is coming but noting that they can’t do that without being able to verify that their competitors were doing the same. That’s why it’s critically important that we build the tech needed to verify AI agreements. On this week’s episode of Your Undivided Attention, I sat down with two experts on AI governance, @fiiiiiist and @janet_e_egan, to talk about the kinds of verification technology we need for AI, the challenges of building it, and the world it could unlock if we did. Given how critical verification technology is, you would assume there are thousands of people working in it. In reality, there’s only around fifty. We urgently need to put our attention, time, and energy into this critical area. Check out our full conversation: t.co/FimnY84Plw
Help research and identify solutions to the ‘intelligence curse’ and gradual disempowerment from AI! ⬇️ x.com/luke_drago_/st…
Important area of near universal agreement. No one should be able to order a bioweapon through the mail. I am a proud signatory of this open letter. x.com/AlecStapp/stat…
Incredible work from @FLI_org on articulating a better "pro-human" path for AI, and the steps that various stakeholders can take to help us get there: betterpathfor.ai x.com/AnthonyNAguirr…
Important AI redline being crossed with new hacking x self-replication research from @PalisadeAI — stemming from significant increases in both of these AI capabilities over the last year. This is the kind of evidence, along with recent evidence of Mythos-level hacking, x.com/PalisadeAI/sta…
"AI will cure cancer" is one of the most potent, seductive arguments in the race to superintelligence. I lost my mother to cancer in 2018. When someone tells me a technology might have saved her — might save someone else's mother next year — every part of me wants to believe it.
Exciting news! @theaidocfilm is now available to rent or buy on Apple TV, Prime Video, or wherever you get your movies on demand. If you missed the theatrical release, you can now watch this urgent film at home. Invite your friends, your family, your church group, your business
Apple TV: tv.apple.com/us/movie/the-a…
This week on Your Undivided Attention, we're bringing you a recent conversation @aza and I had with @Oprah on The Oprah Podcast, where we went deep into the anti-human future that the current AI path is moving us toward. This was an audience of about 150 people who saw a
Watch - bit.ly/41HGPk1
From early employee at Facebook @rosenstein and co-inventor of “Like” button affirming the core problem presented in “The AI Doc” about the reckless race driving an anti-human future. Justin offers clear recommendations on how we need more democratic control, ownership and x.com/rosenstein/sta…
