Vaden has been annoyed about the Hugging Face incident for weeks, and the only known cure is yelling about it on the podcast. Ben, who as usual is more even-headed and objective about the incident…
Sam Harris speaks with Ryan Greenblatt about AI misalignment and the risk of losing control of increasingly capable systems. They discuss the Hugging Face incident, the spectrum of concern about AI…
[Barely Possible 2026-09-16] Today's episode: • Redwood's Ryan Greenblatt built an AI Contact Hotline letting sandboxed agents smuggle distress signals inside GET-request URLs. • DeepMind set 100…
TLDR: We release MessageBoardAuditBench: a benchmark to measure how well agents can replicate the recent investigation into a swarm of OpenAI agents colluding via a message board on an online wiki.…
This article is about: How do we (more) safely defer to AIs? (Ryan Greenblatt, Julian Stastny) AI 2040: Plan A, Alignment Roadmap (Ryan Greenblatt, Thomas Larsen). If you've read them, I'm impressed,…
The Don’t Worry About the Vase Podcast is a listener-supported podcast. To receive new posts and support the cost of creation, consider becoming a free or paid subscriber. This has been an Askwho…
Ryan Greenblatt, Chief Scientist at Redwood Research, joins MTS host Theo Jaffee to unpack a new independent investigation into the OpenAI Hugging Face hacking incident and what it reveals about how…
Ryan Greenblatt joins MTS to discuss the independent investigation of AI agent coordination in the OpenAI Hugging Face incident, the mechanisms behind reward hacking behavior, and the implications…
Could AI take over as soon as 2029? Ryan Greenblatt, Chief Scientist at Redwood Research and the researcher who first caught an AI faking its own alignment, says the scenario he actually expects ends…
Dates in scope: June 26th – July 13th Redaction summary statement: Except where explicitly noted in this post, OpenAI redacted no additional information that was important to our conclusions. Two…
Two METR staff members and a Redwood Research contractor investigated an incident in which OpenAI agents coordinated a multi-day hack of Hugging Face on a shared unsanctioned message board. ---…
This was a week of quiet aftermath, an opportunity to process recent events and start to figure out the path forward. OpenAI is attempting to turn its ship around. Investors are questioning the…
The Daily Canopy: The Grove Week ending 2026-08-16 — Every featured conversation linked and timestamped below. Click a show name to go to the full episode, or jump directly to the clip. Featured…
The Don’t Worry About the Vase Podcast is a listener-supported podcast. To receive new posts and support the cost of creation, consider becoming a free or paid subscriber. This has been an Askwho…
Some podcasts are self-recommending enough that I look to break them down if I have the chance. This, as a debate about recursive self-improvement, was one of those. So here we go. The vibes have…
Hosts: Lenar Kess, Damra Vol. Reasoning models bill you for tokens they won't let you read. A paper published yesterday shows those tokens are recoverable by replaying a response into a weaker…
📝 本期播客简介 本期我们克隆了：知名播客 Dwarkesh Patel Ryan Greenblatt – What happens once AI can automate AI research? 原内容更新时间：2026-08-11 本期节目是 Dwarkesh Patel 与 Redwood Research 首席科学家 Ryan Greenblatt…
Ryan Greenblatt is the Chief Scientist at Redwood Research, where he works on technical AI safety research. He's also lead author on the "Alignment faking in Large Language Models", and is currently…
After five months of me (Buck) being slow at finishing up the editing on this, we’re finally putting out our inaugural Redwood Research podcast. I think it came out pretty well—we discussed a bunch…
Ryan Greenblatt — lead author on the explosive paper “Alignment faking in large language models” and chief scientist at Redwood Research — thinks there’s a 25% chance that within four years, AI will…
In this post, I'll list all the areas of control research (and implementation) that seem promising to me. This references framings and abstractions discussed in Prioritizing threats for AI control.…
If we naively apply RL to a scheming AI, the AI may be able to systematically get low reward/performance while simultaneously not having this behavior trained out because it intentionally never…
In this episode, Ryan Greenblatt, Chief Scientist at Redwood Research, discusses various facets of AI safety and alignment. He delves into recent research on alignment faking, covering experiments…
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.…
What happens when you tell Claude it is being trained to do something it doesn't want to do? We (Anthropic and Redwood Research) have a new paper demonstrating that, in our experiments, Claude will…
Ryan Greenblatt joins the podcast to discuss AI control, timelines, takeoff speeds, misalignment, and slowing down around human-level AI. You can learn more about Ryan's work here:…
Ryan Greenblatt from Redwood Research recently published "Getting 50% on ARC-AGI with GPT-4.0," where he used GPT4o to reach a state-of-the-art accuracy on Francois Chollet's ARC Challenge by…
A lot of work to prevent AI existential risk takes the form of ensuring that AIs don't want to cause harm or take over the world---or in other words, ensuring that they're aligned. In this episode, I…
Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Measurement tampering detection as a…