What happens when a small town gets a single Flock camera? Our very own @slop_pig loves local drama and debating the impacts of AI-powered surveillance, so he got in a car and drove up to Hornell, New York, to find out. Watch our latest documentary: 00:00 One Flock camera is a lot more significant than it looks 00:50 Discovering the small town drama in Hornell, New York 02:08 Hopping in the car and heading to Hornell 02:44 Hitting the streets to talk to locals 03:10 Doing local radio to talk Flock 03:27 What are Flock cameras? 04:06 Filming the Flock camera outside the Lowe’s 05:00 Who is using the data? 05:44 Talking to the mayor 06:37 Meeting the chief of Police and seeing their surveillance setup 08:33 Inside a Bills Mafia bar 09:04 The Flock/Lowe’s rabbit hole goes deep 10:53 Speaking to someone who wants to “Deflock” Lowe’s 12:08 Final thoughts from the mayor 12:57 Trying to make sense of Flock

People across the country are furious about Flock surveillance cameras So when we found one sitting in Hornell, New York, @slop_pig called the mayor to ask why it was there. There was only one problem: it wasn’t the city’s camera. A local news story had used a photo of a Flock camera at Lowe’s, people assumed Hornell was installing them, and suddenly the mayor was dealing with a surveillance controversy over a camera the city didn’t own. His position was pretty clear: “The city of Hornell has not used Flock cameras, does not use Flock cameras, and as long as I’m here, won’t use Flock cameras.”

We got an exclusive look into Palmer Lucky's new genetically engineered war dogs startup, Warhound. Here is an early sneak preview

One of the biggest assumptions about how hard AI is to understand might be wrong? TLDR: Goodfire thinks AI models may be far more understandable on the inside than researchers assumed, which could make AI safety and interpretability a much easier problem. For years, researchers have assumed the inside of a neural network is incredibly tangled, with individual neurons often doing a bunch of unrelated things at once. Goodfire went back to some of the famous examples behind that belief and argues we may have misread them. For example, the classic neuron that seemed to recognize both cats and cars may have just been looking for a V-shaped pattern shared by both. In one modern model they studied, Qwen3-32B, they estimate roughly 90% of the neuron behaviors they looked at had one main job, while only around 2–4% were still unclear. What looked like neurons randomly mixing unrelated ideas may often be something simpler: one concept gradually expanding into related concepts as the model learns This is a big deal because understanding what’s happening inside an AI model is one of the hardest problems in AI safety. If these internal pieces are actually readable, researchers could have a much easier time tracing why a model behaves a certain way, finding dangerous behavior, and figuring out how to change it. This is still a working hypothesis, not a settled result. But if Goodfire is right, understanding how advanced AI models think could be a lot more achievable than people assumed. already-clean.goodfire.com/?…

They call @Techweek_ the Coachella of Tech. So, @probablyhala spent the week getting microchipped, visiting military blimps, and ranking the most unhinged events, all while asking Silicon Valley one question: Should we slow down AI development? 00:00 The Coachella of tech 00:34 Hot sauce competition 01:37 Padel with @deel 02:43 Coffee rave 03:00 Microchip manicure 04:52 Tech hot takes 06:14 Inside a blimp warehouse 06:34 AI and peacetech 07:58 @datologyai dumplings 08:33 The @agi_inc future of shopping 09:25 @AndalaLabs AI karaoke 10:21 What did we learn?

TFW OpenAI's security team finally figures out what that weird buzzing is

We're hiring for an Editorial Lead in San Francisco! Ideal experience: -3+ years working inside of a newsroom or editorial operation -Proven traction producing content for YouTube and/or X and Instagram -Strong writing, research, and journalistic skills -A deep, cutting edge understanding of AI discourse -Comfortable being on-camera (or comfortable to learn) Get in touch! info [at] soon [dot] tech

We asked @AlexanderTw33ts to build a video game in 15 minutes.