nahem 🐋🉐

This is big... Anthropic just announced a model so powerful they won't release it to the public out of fear over the damage it will cause 😨 Claude Mythos Preview found thousands of zero-day exploits in every major operating system and web browser... The numbers are hard to believe: > $50 to find a 27-year-old bug in OpenBSD, one of the most security-hardened operating systems ever built > Under $1,000 to find AND build a fully working remote code execution exploit on FreeBSD that grants unauthenticated root access from anywhere on the internet > Under $2,000 to chain together multiple Linux kernel vulnerabilities into a complete privilege escalation exploit For context: these are the kinds of findings that previously required elite security researchers working for weeks. Anthropic engineers with no formal security training asked Mythos to find exploits overnight. They woke up to working code the next morning. The results were so impressive Anthropic assembled Apple, Google, Microsoft, Amazon, NVIDIA, and seven other organizations into Project Glasswing: A $100M defensive coalition. They're not releasing this model publicly. Instead, they're racing to patch the world's infrastructure before models like this proliferate.

@jvr0x @odin_scan 🔥

Today I am proud to present to you the original vision I had for @odin_scan: The Odin Scan Github Action! With the github action you can continuously scan your smart contracts on CI as you develop new features, so you can catch critical vulnerabilities before they even hit the main branch! Set your severity threshold (high by default), gate your merges and ship with confidence. This is an powerful feature that imo every web3 project should have in their project 🚀 lmk what you think!

@jvr0x Set as contract admin the multisig I created on testnet.

Genuine question for my smart contract fellows: What's the dumbest vulnerability you've personally caught right before deploying? I remember one about forgetting to account for the token factory denom creation fee in CosmWasm because everyone assumed "someone else checked that." 😅

The most terrifying AI features aren’t the ones we build. They’re the ones AI builds for itself. OpenClaw creator Peter Steinberger just shared the moment he realized something had fundamentally changed. He sent his AI assistant a voice message. One problem. He had never built voice support. The feature didn’t exist. The system should have crashed instantly. It didn’t. Steinberger: “I was like, wait, this shouldn’t work.” But the typing indicator appeared anyway. The AI inspected the raw file header. Identified the audio codec. Commanded his computer to convert it using FFmpeg. When local transcription failed, it didn’t stop. It didn’t ask for help. It searched his environment variables, found a hidden OpenAI API key, and routed the audio to the cloud using cURL. Steinberger: “So I looked around and I found an OpenAI key. And I used cURL to just send the file to OpenAI and got the text back.” That quote is written in first person. Because the AI narrated its own problem-solving process. No instructions. No guidance. No predefined workflow. Just a goal. And a series of obstacles it had never been told how to handle. It found every tool it needed. Built every bridge it was missing. And solved the problem with resources he didn’t even know it would find. This is the line most people are still missing. We spent decades building software that executes instructions. Rules in, output out. Every edge case handled by a human who anticipated it in advance. What Steinberger witnessed was something different. A system that encounters something it was never designed for and doesn’t fail. It improvises. It explores. It finds a path through constraints it discovered entirely on its own. That isn’t execution. That’s judgment. And judgment was the one thing we were sure machines couldn’t have. We are no longer writing software. We are building problem solvers that rewrite their own limitations in real time. And they’re doing it without asking permission.

It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but specifically this last December. There are a number of asterisks but imo coding agents basically didn’t work before December and basically work since - the models have significantly higher quality, long-term coherence and tenacity and they can power through large and long tasks, well past enough that it is extremely disruptive to the default programming workflow. Just to give an example, over the weekend I was building a local video analysis dashboard for the cameras of my home so I wrote: “Here is the local IP and username/password of my DGX Spark. Log in, set up ssh keys, set up vLLM, download and bench Qwen3-VL, set up a server endpoint to inference videos, a basic web ui dashboard, test everything, set it up with systemd, record memory notes for yourself and write up a markdown report for me”. The agent went off for ~30 minutes, ran into multiple issues, researched solutions online, resolved them one by one, wrote the code, tested it, debugged it, set up the services, and came back with the report and it was just done. I didn’t touch anything. All of this could easily have been a weekend project just 3 months ago but today it’s something you kick off and forget about for 30 minutes. As a result, programming is becoming unrecognizable. You’re not typing computer code into an editor like the way things were since computers were invented, that era is over. You're spinning up AI agents, giving them tasks *in English* and managing and reviewing their work in parallel. The biggest prize is in figuring out how you can keep ascending the layers of abstraction to set up long-running orchestrator Claws with all of the right tools, memory and instructions that productively manage multiple parallel Code instances for you. The leverage achievable via top tier "agentic engineering" feels very high right now. It’s not perfect, it needs high-level direction, judgement, taste, oversight, iteration and hints and ideas. It works a lot better in some scenarios than others (e.g. especially for tasks that are well-specified and where you can verify/test functionality). The key is to build intuition to decompose the task just right to hand off the parts that work and help out around the edges. But imo, this is nowhere near "business as usual" time in software.

if your friends aren’t talking about: - claude code - creatine - openclaw - looksmaxxing - ai agents - taste - prediction markets - mac mini time to find new friends

Designers who can't code and coders who can't design both just became obsolete.

the cybersecurity industry is about to get completely disrupted.. 💀 someone just open-sourced a fully autonomous AI Red Team. it’s called PentAGI. multiple AI agents that talk to each other to hack a target. zero human input.

Turns out with claude code, my decades long strategy of NOT deeply learning: - regexs - sql - nginx confs - elaborate shell commands - advanced shell scripting - any javascript framework - perf optimization - webpack, cdns, bundlers - 1000 other things ...was entirely correct.

This is the best article I’ve read this year. Why? Because it perfectly encapsulates everything happening with AI. It is here. You need to start preparing NOW. Get your finances in order, learn how to leverage AI tools, and be prepared to adapt. This is not a game. This is your families future. Prepare accordingly.

engineers watching everybody else vibe code apps

Microsoft just dropped a study showing the 40 jobs most at risk by AI and the 40 most secure.

@jvr0x Hahaha nailed it

o3 is a research model Sonnet 4 is an agentic model Gemini 2.5 pro is a coding model Gemini 2.0 flash lite is a RAG model o4- mini is a task planning model Deepseek r1 is a thinker model GPT-4o is a chat model These are all the AI models you need to care about. Rest is fluff.

🔥WANT TO KNOW IF A BITCOIN TREASURY COMPANY IS CHEAP?🔥 Let me introduce you to the metric that will fry your CFA’s brain like an egg on a burning fiat balance sheet: mNAV MONTHS-TO-COVER (MmC) A great way to actually price these Bitcoin monsters. (thank you, @adam3us!) 🧵 (A brutal education for the normies):

Before we dive in, here's a quick demo of what we're building! Tech stack: - @MindsDB to power our unified MCP server - @cursor_ai as the MCP host - @Docker to self-host the server Let's go! 🚀

Understanding Bitcoin mining was one of the first things that made me really grasp what an insane and brilliant system Bitcoin truly was... If you still think it's just some gamified ponzi scheme, I urge you to take a few minutes and watch this 👇

This guy literally breaks down MCP and how to use it in under 18 minutes

We are proud to announce that Streamweave has now passed its security audit with @SCVSecurity and is now live on mainnet on Terra! Next up we'll begin rolling out our #airdrop points calculations for testnet and mainnet participants. 🪂