This day is still open for picks. It closes at midnight UTC.
New Anthropic research: progress on provable inference for large language models. Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so. These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers. For our full write-up, see here: anthropic.com/research/prova… Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: github.com/anthropics/Fla… We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (rdalal.github.io) and Ben Mayer (@transmissions11).
On Sunday, Adam Robinson got on a private jet in Austin with his wife, their 4-year-old, their 2-year-old and their nanny for a trip to the Bahamas. None of them made it home. He left Wall Street to start a company and got stuck at the same revenue for four years before he finally cracked it. What came after was a $22M-a-year business, built without a dime from investors. The idea behind it sounds like a magic trick. You visit an online store, poke around, never type in your email, and leave. His software could still figure out who you were, so the store could email you anyway. I interviewed Adam for my podcast in January 2025. He gave me two and a half hours, the longest interview he'd done up to that point, and he answered every single thing I asked. He told me how much he paid himself and what he spent buying the Retention.com domain, then walked me through his churn problem and how he planned to fix it. He talked about his own money the way most people talk about what they had…
This might be the greatest issue report in history Some folks wrote a paper about my /retro skill assessing its effectiveness Incredible arxiv.org/pdf/2610.34938
this is basically the bitter lesson at work the smartest models will need fewer and fewer tokens, they won't need to verify their work or check it for bugs, they will just be right
Crazy how Large Language Models have gotten good at pretty much everything other than generating text
Things will change again and again and again. We’re still very early and agents today are incredible, and yet so frustrating still. Can’t wait to crack things to get that flow state back. Long journey ahead.
💸 Update: My total biz expenses have now dropped to just ~$300/mo! This is to run all my websites/apps serving per year: - 20,000 paying customers - 252 million visitors - 2.4 billion requests - 2.4 petabyte (PB) in bandwidth (~$8,000/mo if you add the $7,000/mo GPU bill and $800/mo R2 storage I pay for Photo AI, I exclude it above because it's one project that's so much more expensive in size than all the other sites together, but included in the chart below) Claude Code helped me to find smart ways to save more money, AI is just really really really good at finding cost savings by itself, which is not so great for the economy short-term, but long term it trims all the fat and opens up spending for better investments I think Anyway here's what it saved me in money, almost $34,000/mo: - $29,000/mo - switched to better and cheaper video models on Photo AI, video was very expensive and since I didn't update it ever it stayed expensive, also routed some image jobs to Google Cloud w…
Americans don't use WhatsApp because of a decision Apple made in 2013. We only know about it because the emails came out in court, and the whole thing came down to keeping your kids off Android. It started with Eddy Cue, the guy who runs Apple's services. That April he emailed the team saying Apple really needed to put iMessage on Android. Google was reportedly trying to buy WhatsApp for $1 billion at the time, and Cue's pitch was that Apple already had the best messaging app, so why not make it the one the whole world used. Craig Federighi, Apple's software chief, shut it down. He asked how Apple was supposed to get Android users to leave WhatsApp when they didn't even have iPhone friends. So back in 2013, the guy running iPhone software was basically admitting WhatsApp already owned everybody who didn't have an iPhone. Then he got to what he was actually worried about. iMessage on Android would remove "an obstacle to iPhone families giving their kids Android phones." Phil Schill…
ChatGPT and Codex are helping engineers build for the physical world. We’re sharing our progress across mechanical design, electrical engineering, and simulation—from helping our hardware teams iterate faster to testing new workflows with industry collaborators.
Friday was my last day at Anthropic. Two years ago, when Claude Code was a small internal tool, I asked it to build an RL env. I steered it for tens of hours but never manually wrote a line of code. It worked, I was hooked, and I joined the team. When we first launched, most people approved every permission prompt by hand and mostly used Claude to answer questions and write tests. Today, the default is trusting Claude's judgment to auto-approve. Claude does large migrations and builds new products. Helping build Claude Code into an industry-defining product has been the journey of a lifetime. Thank you to the incredible team behind Claude Code, Cowork, and Tag. These products are the sum of thousands of great ideas from you. I'll miss working with you all. Thank you to our users for embracing a new way of working. Your constant feedback shaped so much of what we shipped and your stories made the work so meaningful. Excited to see what you all make next!
