Slightly reducing the sloppiness of AI generated front end
Slightly reducing the sloppiness of AI generated frontend Back to Blog | RSS | Atom 2026-06-12 I’ve been trying to make progress on a herculean task; I want decent looking programs for personal use which I can generate quickly using AI agents yet I’m a person without taste controlling an AI without taste. I found […]
A Call to Action: Stop the FCC’s KYC Regime
Robocalls are really annoying. Everyone knows the misery of scam calls, spoofed numbers, fake warranty pitches, fraudulent bank alerts, and automated political spam. The FCC is correct to claim that illegal calls erode trust in the phone system and cost Americans time, money, and security. But this problem does not justify a dragnet solution. Under […]
How to automate Instagram engagements with computer vision (and get banned)
Obviously, Instagram does not want you to automate engagement. Their HTML is a mess of randomly generated class names and deeply nested divs. The structure changes every deployment. Any script that relies on DOM selectors breaks within weeks because the class name doesn’t exist anymore. But it doesn’t matter anyway. Instagram can obfuscate their code […]
WASI 0.3.0 Released
WASI 0.3 is official, and async is now native to WebAssembly Components. The WASI Subgroup voted to ratify WASI 0.3.0, rebasing WASI onto the WebAssembly Component Model’s async primitives. Most of the changes in the 0.3 interfaces are entirely mechanical. WASI 0.2 had to perform some acrobatics to make async work, but now that async […]
Hazel (YC W24) Is Hiring a Full Stack Engineer
Note: This role is exclusively available to U.S. citizens who can obtain and maintain a U.S. government-issued Top Secret security clearance with full-scope polygraph. There are no exceptions. Who We Are: Our mission at Hazel is simple and powerful. We use AI to help U.S. government purchasing teams buy at the speed of need with […]
Maxproof
Abstract:We present MaxProof, a population-level test-time scaling framework for competition-level mathematical proof in the MiniMax-M3 series. M3 first trains three proof-oriented capabilities — proof generation, proof verification, and critique-conditioned proof repair — using a defense-in-depth generative verifier engineered for low false-positive rate. These capabilities are merged into a single released M3 model. At test time, […]
Ryanair dark UX patterns summer 2026 refresher
Everyone likes dark UX patterns – such fun! Ryanair are Europe’s most profitable airline and they are masters of this noble form. This is an all time classic from around 8 years ago – to not buy travel insurance, you must select Don’t Insure Me, midway down a list of countries:
Kimi K2.7-Code: open-source coding model with better token efficiency
1. Model Introduction Kimi K2.7 Code is a coding-focused agentic model built upon Kimi K2.6. With substantial improvements on real-world long-horizon coding tasks, it strengthens end-to-end task completion across complex software engineering workflows while improving token efficiency, reducing thinking-token usage by approximately 30% compared with Kimi K2.6. 2. Model Summary Architecture Mixture-of-Experts (MoE) Total Parameters […]
The Future of Email
Email authentication: the trust layer that the future of email depends on Email has always had a spoofing problem. Anyone can put anything in the “From” field of an email. For most of email’s history, that was manageable. A careful reader could catch the tells, such as a slightly off domain name, implausible urgency, or […]
Report on an Unidentified Space Station
Report on an Unidentified Space Station Survey Report 1 By good luck we have been able to make an emergency landing on this uninhabited space station. There have been no casualties. We all count ourselves fortunate to have found safe haven at a moment when the expedition was clearly set on disaster. The station carries […]