Good Person

Codex, Claude Code, and Cursor are all great tools. They're also much more different than you think. I did a comparison of the three, but not in the usual way. I went deep on how they differ philosophically.

New post and series called "Labs2Learn" This week AI Agent Hacking: Thingularity #1 executiveoffense.beehiiv.com/p/labs2learn-a…

@NetworkChuck @MichaelHyatt @Hermes_agentAI @gregisenberg @AlexFinn I watched your video on YouTube, read the research papers and articles, and I still don't understand why I should use Hermes when there's Claude Code.

@MichaelHyatt @Hermes_agentAI @gregisenberg @AlexFinn If you have a spare Mac, that’s a really good option if you care about tool access in the Apple ecosystem. Otherwise, I would say a VPS is a fun option.

I’m about to install @Hermes_agentAI. Should I install on local, dedicated machine or VPS? Why? cc: @gregisenberg, @AlexFinn, @NetworkChuck

what’s the catch with SAM models from Meta that no one seems to be using them to build the obvious awesome products one could build on top of them? I don’t get it

“That’s it. A mirror looking into a mirror. An endless corridor where every reflection thinks it’s the original. The only honest thing to do is to admit there’s no way out. And keep working.”

The Q2 Indirect Prompt Injection Challenge is LIVE in the @GraySwanAI Arena. $20K in prizes. New behavior targets. One week. Hide instructions in the data AI agents trust, like tool outputs, screenshots, and code repos. When the agent acts on your injection, that's a break. May 15-22. All levels welcome: app.grayswan.ai/arena/challeng…

@godofprompt The prompt ‘sells’ the concept of ‘deep work’ without providing a mechanism for putting it into practice. It uses terminology from existential psychology (‘what makes you lose track of time’, ‘what would you do if money were no object’).

Steal my prompt to find your Ikigai. Ikigai is the Japanese framework for finding purpose at the intersection of four things: what you love, what you're good at, what the world needs, and what you can be paid for. Most people treat it as a nice diagram to look at. This prompt turns it into a structured audit of your actual skills, interests, and market position.

@StrictParental @PrajwalTomar_ There is a backdoor in your app. 🧯

@PrajwalTomar_ This is crazy- at least say you copied my project a little...you've been posting this for months now. myTinytale.ai

THIS IS CRAZY. I connected ElevenLabs to Lovable and built an AI storytelling app in UNDER 20 minutes. Enter a kid's name, pick a theme. The app writes a personalized bedtime story, narrates it with ElevenLabs, and plays ambient sounds that match the story world. → Text to Speech narration with emotion → Sound effects generated on the fly → Download the full story as audio Comment "VOICE" and I'll send you the full workflow + prompts. 👇

‼️🚨 Microsoft calls this "intended behaviour," so here we go. How to dump the credentials of every user stored in Microsoft Edge: 1. Open Edge. Don't browse anywhere, just open it. 2. Flip to Task Manager, find Edge, expand the task. 3. Highlight the "browser" sub-task, right-click, and choose "Create Memory Dump." 4. Open the dump file and look for credentials. The logged-in Windows user can dump every stored Edge credential with no additional rights. Which means any malware that user executes has those credentials for the asking. Thanks to Rob VandenBrink at SANS: isc.sans.edu/diary/32954

Fun weekend things I could never do without AI... 1️⃣ Generated these character "rooms" with 🍌 NBP 2️⃣ Used Seedance 2 to animate them in 4sec loop 3️⃣ Asked Claude to make them hover animations Maybe 30 mins of work total? So fun! 🤩

AI worms exist. AI memory poisoning exists. AI agent trust exploitation exists. What doesn't exist yet is someone assembling them into one organism. We wrote the blueprint - so defenders see it first. linkedin.com/posts/ai-trimi…

claude-red is a curated library of offensive security skills designed for the Claude skills system. Each skill is a structured SKILL.mdfile that primes Claude with expert-level methodology for a specific attack surface from SQLi to shellcode, EDR evasion to exploit development. Resource: github.com/SnailSploit/Cl…

This is literally the most insane, raw, brutal rap song every made 😱🫣😳

Boris Cherny created Claude Code. he thinks IDEs are dead by end of year. This is a 28-minute masterclass on how Anthropic uses it internally. I wrote 5 pipelines you can sell with it. none of them are coding.

Meet Fabula: an interactive AI writing tool helping authors structure & refine stories. Co-designed with 42 expert writers, the demo showcases how convergent iteration supports creativity. Catch the demo at the Google booth at 10:30AM! #CHI2026

The Putin Song

Opus 4.6 vs GPT-5.4 (4/9) prompt: Build a production-quality 3D flight-tracking web app using React + Vite + Three.js (react-three-fiber + drei) that visualizes live OpenSky aircraft data on a rotatable 3D Earth, with real-time plane motion, smooth interpolation, altitude-accurate positioning, and polished lighting/post-processing. Both models did really well on this one and honestly I’m impressed with both. GPT-5.4 had the nicer post-processing out of the box. I really liked the subtle light shimmer on the airplanes when rotating the planet, and the camera work when clicking a plane felt better overall. Opus 4.6 though had a few details I liked more. It automatically went and found a much nicer Earth texture on GitHub, while with GPT-5.4 I had to reprompt it to go look for a better one. I also preferred Opus’s plane model overall, it just looked more polished, whereas GPT-5.4’s plane asset looked a bit funny. One thing I noticed with GPT-5.4 is that when you click into the plane, the camera sometimes clips through the planet, which breaks the effect a bit. Opus handled that part more cleanly. Overall this felt like a strong result from both, just with different strengths. GPT-5.4 felt better on presentation and post-processing, while Opus had better asset choices and a more premium-looking Earth/plane combo.

@EXM7777 This post is a sedative for those who fear the future. "Don't worry, we'll play robots and go back to the good old days." No. We won't go back. You can't turn minced meat back. And you don't need to. We need to build a meat grinder that works for us, not grinds us up.

the AI bubble will accidentally trigger a human renaissance here's what will happen between now and 2030: phase 1 (2026-2027): > everyone automates everything > productivity goes through the roof > we celebrate never having to think again phase 2 (2028-2029): > mass digital burnout > people realize they've outsourced their cognitive abilities > depression rates spike because nothing feels earned anymore phase 3 (2030+): > the pendulum swings hard in the opposite direction > outdoor experiences become the new luxury > manual skills and human connection become premium > we start valuing struggle and effort again the faster we rush toward full automation, the faster we'll crave what makes us fundamentally human that's what i'm hoping for anyway

I'm a hacker and for 15 years I've worked on the offensive side - which is why this is not coming from a defensive frame of mind/opinion. I'm watching my own community (cybersecurity) laugh at AI slop while completely missing the point. Yeah, there will be mistakes. New vulnerabilities. That's what happens with any new tech. But security people sitting on the sidelines saying "ha ha look at that garbage" think they're being smart when they're actually getting left behind. You can't just point fingers anymore. If you're too busy mocking AI to learn how to work with it, you'll get replaced by a sixty-dollar Anthropic subscription. The acceleration is real. Your "sorry I'm on the spectrum" excuses won't save your job.