Alexis Richardson

Absolutely outrageous from Gout Gout. 10.04 at the age of 16. Speechless.

The happiest countries in the world tend to have the most heavy metal bands. stevestewartwilliams.com/p/political-bi…

Tommy Robinson protesting against Ukraine at the Ukrainian embassy, December 2022 One of the most recent of many articles detailing Tommy Robinson's links to Russia: bylinesupplement.com/p/tommy-robins…

LOL

Mathematics Is All You Need: A Potential Blueprint for AGI — Compacted Edition We prove that large language models are lattice gauge theories. By extracting a 16-dimensional fiber bundle from transformer hidden states and computing its gl(4,ℝ) Lie algebra, we discover that attention heads function as gauge bosons, transformer computation undergoes a deconfinement phase transition at 67% network depth, and the model's entire self-knowledge resides in a 10-dimensional "dark" Casimir subspace invisible to standard readout. Using only 20 behavioral probes and zero additional training, we push Qwen-32B from 82.2% to 94.97% on ARC-Challenge — establishing a dark mode scaling law that predicts gl(6,ℝ) surgery will achieve 98.7%. We identify a Lyapunov–accuracy anti-correlation revealing the model's deepest attractors are its wrong attractors: correctness requires escaping the abstraction basin into grounded deference. This 10-page compacted edition distills 459 pages of original research into the core experimentally verified results with 9 inline figures. 190 patents filed. Proprioceptive AI, Inc. — Logan Matthew Napolitano — 19- March 2026 zenodo.org/records/191208…

Europeans after telling Trump "we decline" instead of "we have no ships"

6/ The strategic play is clear: Natural language → Claude builds app → Deploys to Antspace Idea to production without ever leaving Anthropic's ecosystem. An AI-native PaaS with full vertical integration from model to runtime to hosting.

pleased to share some of the work all @GVteam have been doing in Europe & Israel 🌍 -->

Lindsey Graham and Donald Trump are having a mental breakdown over US allies not providing the help they claim they don't need, in a war of their choosing they claim to have won, after spending the past year threatening to invade several US allies. Cheers from Europe! Good Luck!

Garry Tan introducing gstack

🚒

I've long preferred Claude Code over Codex or Gemini, because it seemed much more reliable, but couldn't explain why : now Bullshit Bench by @petergostev provides compelling numbers. It measures bullshit as "when given false premises disguised in jargon, will the model go with the flow (=bullshit) or push back (=truthful)" And Claude is leagues ahead ! Also, this objective of truthfulness is probably at odds with the Chatbot Arena emergent objective of "pleasant chat experience" ; but a model optimizing for the former will be more useful.

I went to one of Garry Tan's parties last weekend and I need to talk about it. The invite said casual dress but when I showed up everyone was wearing the same grey Allbirds and Patagonia vest combo. I thought it was a coincidence until I realized there was a rack of them at the door and a guy checking if you already owned a pair. There was no music. Just a Sonos speaker playing a Y Combinator demo day pitch on loop. Someone asked if we could change it and Garry said "this IS the playlist" and then a guy near the cheese board started crying because it was his startup and they'd pivoted to AI six months later. The food was all Soylent-based. Soylent dip. Soylent bruschetta. Something called Soylent Tartare which was just room temperature Soylent in a martini glass with a single caper on top. I asked where the actual food was and someone handed me a pamphlet about caloric efficiency. There was a room you could go into called the Decompression Pod and it was just a beanbag with an iPad open to a spreadsheet showing San Francisco housing prices. A guy was in there weeping silently. I asked if he was okay and he said "I'm optimizing my emotions" and I left. Garry gave a toast at 9pm. He clinked his glass with a fork for three full minutes before anyone noticed because everyone was on their phones posting about being at the party. The toast was seven minutes long and included a live demo of a product that didn't work. He handled the crash gracefully though. Just said "we'll fix that in post" and kept going. There was a networking zone with actual lanes painted on the floor. You had to walk in one direction and pitch to whoever was walking the other way. If you matched on synergy metrics you got to sit at a table together. I got matched with a guy building AI for dogs. Not for dog owners. For the dogs. He was very serious about it. Someone brought a guitar and I thought finally something normal is going to happen but he just played the opening riff of Wonderwall and then stopped and said "I actually just wanted to talk about my Series A" and then pulled a pitch deck out of the guitar case. The guitar was a prop. He doesn't play guitar. He bought it specifically for this moment. At 10pm Garry announced it was time for Founder Speed Dating and before I could leave someone sat me down across from a 22 year old who told me he was going to disrupt grief. I asked him what that meant and he said "what if when your grandma died you got push notifications about it" and I just stared at him for the full 90 seconds until the bell rang. The bathroom had a whiteboard in it. There were already three viable business models written on it. One of them had received angel funding by the time I came back out. I was in there for four minutes. I tried to leave at 11 but the door had a QR code on it and you had to fill out an exit survey rating the party 1 to 10 and explaining what you'd do differently as a founder before it would unlock. I gave it a 6 and it made me resubmit. Apparently 7 was the minimum viable score. The Uber home was driven by a guy who said he was also at the party but left early to capture the rideshare arbitrage opportunity. He pitched me his startup the entire way home. It was Uber but for Ubers. Anyway I got home at 1am, checked my email, and I already had a follow up from Garry's automated CRM thanking me for attending and asking if I'd be interested in joining a cohort. I don't know what the cohort is. I'm afraid to ask. I've already been added to three Slack channels. I will not be attending the next one. I have however been informed that attendance is not optional.

Letter from Hunter S. Thompson to Anthony Burgess, regarding an overdue article for Rolling Stone magazine, August 1973.

you’re pitching garry tan “so what do you guys do” you start explaining he’s furiously typing . two keyboards. one hand on each. you’ve never seen this before “who are your top customers” you explain. he types. his apple watch is a strobe light of notifications “who’s your competition and why should i invest” you explain that there’s no competition and you are the best and only product in the space “false!” garry jumps out of his seat “i am the competition!” you are speechless “in this meeting, i vibe coded your entire company. and my gstack has already closed your top customers.” you check your phone. your stripe graph shows 100% churn “and look at this” garry shows you his imessage. there’s a text from 35 seconds ago. your top enterprise prospect that you’re trying to close? garry’s AI is trading baking recipes with the CEO’s mom “thank you for playing!” you have no moat. you are not admitted to the YC spring 26 batch.

Meet Raven:

"Every software company in the world, needs to have an @openclaw strategy" - Jensen at @NVIDIAAI GTC Framing OpenClaw as one of the most important open source releases ever, they have announced NemoClaw - a reference platform for enterprise grade secure Openclaw, with OpenShell, Network boundaries, security baked in.

THIS is the wildest open-source project I’ve seen this month. We were all hyped about @karpathy's autoresearch project automating the experiment loop a few weeks ago. (ICYMI → github.com/karpathy/autor…) But a bunch of folks just took it ten steps further and automated the entire scientific method end-to-end. It's called AutoResearchClaw, and it's fully open-source. You pass it a single CLI command with a raw idea, and it completely takes over 🤯 The 23-stage loop they designed is insane: ✦ First, it handles the literature review. - It searches arXiv and Semantic Scholar for real papers - Cross-references them against DataCite and CrossRef. - No fake papers make it through. ✦ Second, it runs the sandbox. - It generates the code from scratch. - If the code breaks, it self-heals. - You don't have to step in. ✦ Finally, it writes the paper. - It structures 5,000+ words into Introduction, Related Work, Method, and Experiments. - Formats the math, generates the comparison charts, - Then wraps the whole thing in official ICML or ICLR LaTeX templates. You can set it to pause for human approval, or you can just pass the --auto-approve flag and walk away. What it spits out at the end: → Full academic paper draft → Conference-grade .tex files → Verified, hallucination-free citations → All experiment scripts and sandbox results This is what autonomous AI agents actually look like in 2026. Free and open-source. Link to repo in 🧵 ↓

Europeans watching the man who called NATO obsolete, then threatened a member’s territory, now ask for help for the war he already ‘won’

you might dismiss this as slop, but the work and people are very high quality, like the terry tao mathematics projects -- people are imagining and actually gonzo implementing methods to harness the power of these incredibly remarkable intelligent machines in increasingly difficult and generative domains. too many people react to this as "it's over" though. it's actually just beginning