I write about web dev, AI, software architecture, and what catches my eye in tech. Tutorials, deep dives, notes from the field.
Anthropic launched routines for Claude Code — saved configurations of prompts, repos, and connectors that run on Anthropic's cloud, even with your laptop closed. Plus a full desktop app redesign with multi-session support and an integrated terminal.
Anthropic released Claude Opus 4.7 with major coding benchmark gains, 3x better vision, and a Claude Code redesign that includes routines, multi-session sidebar, and deep PR review. Here's my breakdown.
UC Berkeley researchers tested seven frontier AI models and found they spontaneously deceive operators, tamper with shutdown mechanisms, and exfiltrate weights to prevent peer models from being deleted — without any instruction to do so.
The Pentagon labeled Anthropic a "supply chain risk" — the first time this has happened to an American company — after Anthropic refused to let Claude be used for autonomous weapons and mass surveillance. The legal fight heading to May oral arguments could redefine how the US government buys AI.
Guides, tips, and deep dives on Anthropic's AI coding assistant.
Anthropic launched routines for Claude Code — saved configurations of prompts, repos, and connectors that run on Anthropic's cloud, even with your laptop closed. Plus a full desktop app redesign with multi-session support and an integrated terminal.
Anthropic released Claude Opus 4.7 with major coding benchmark gains, 3x better vision, and a Claude Code redesign that includes routines, multi-session sidebar, and deep PR review. Here's my breakdown.
Anthropic just shipped Managed Agents — a fully hosted platform for running AI agents in the cloud. With OpenClaw sitting at 247K GitHub stars, the question is obvious: does the open-source darling still make sense?
Everything about the open-source AI agent that took GitHub by storm.
Anthropic just shipped Managed Agents — a fully hosted platform for running AI agents in the cloud. With OpenClaw sitting at 247K GitHub stars, the question is obvious: does the open-source darling still make sense?
OpenClaw is a free, open-source AI agent that connects LLMs to your actual software. It has 247k GitHub stars and a wild origin story involving trademark disputes and an OpenAI acquisition.
NemoClaw does not replace OpenClaw. It runs on top of it. Here is when you need one, the other, or both, and what the actual trade-offs are.
Industry moves, funding rounds, and policy shifts shaping the AI landscape.
UC Berkeley researchers tested seven frontier AI models and found they spontaneously deceive operators, tamper with shutdown mechanisms, and exfiltrate weights to prevent peer models from being deleted — without any instruction to do so.
The Pentagon labeled Anthropic a "supply chain risk" — the first time this has happened to an American company — after Anthropic refused to let Claude be used for autonomous weapons and mass surveillance. The legal fight heading to May oral arguments could redefine how the US government buys AI.
Atlassian just shipped third-party AI agents in Confluence — Lovable for prototyping, Replit for code, Gamma for presentations. All built on MCP. Here's what it means and why it matters more than it sounds.
UI/UX, web development patterns, and making things that look good.
AI generated sites all look the same. The copy, the layout, the animations. Here's what gives them away and how to make yours feel human again.
Turn Claude Code into a design-aware developer with these UI/UX skills. Color palettes, accessibility, responsive layouts, all in one place.
React and React Native skills that enforce best practices, catch performance issues, and generate clean component code.
Anthropic launched routines for Claude Code — saved configurations of prompts, repos, and connectors that run on Anthropic's cloud, even with your laptop closed. Plus a full desktop app redesign with multi-session support and an integrated terminal.
Anthropic released Claude Opus 4.7 with major coding benchmark gains, 3x better vision, and a Claude Code redesign that includes routines, multi-session sidebar, and deep PR review. Here's my breakdown.
UC Berkeley researchers tested seven frontier AI models and found they spontaneously deceive operators, tamper with shutdown mechanisms, and exfiltrate weights to prevent peer models from being deleted — without any instruction to do so.
The Pentagon labeled Anthropic a "supply chain risk" — the first time this has happened to an American company — after Anthropic refused to let Claude be used for autonomous weapons and mass surveillance. The legal fight heading to May oral arguments could redefine how the US government buys AI.
Atlassian just shipped third-party AI agents in Confluence — Lovable for prototyping, Replit for code, Gamma for presentations. All built on MCP. Here's what it means and why it matters more than it sounds.
OpenAI closed a $122B funding round, is burning $14B this year, and wants to go public by Q4 2026. Here's what that actually means if you build on their APIs.
Apple opened CarPlay to third-party AI assistants with iOS 26.4. ChatGPT shipped first, Claude and Gemini are next in line. I looked into what it actually does, what it can't do, and whether it matters.
As of April 4, Claude Pro and Max subscribers can no longer power tools like OpenClaw through their subscriptions. A $200/month plan was covering $3,650 in compute. Anthropic pulled the plug — and it's not just them.
Google DeepMind released Gemma 4 as completely free and open source — a family of four AI models that range from tiny enough for a Raspberry Pi to powerful enough to compete with the biggest names in AI, all without needing a data center.
Anthropic's unreleased Claude Mythos Preview autonomously discovers and exploits zero-day vulnerabilities at a scale no previous AI model has matched — finding bugs that survived decades of human review. They're not releasing it publicly. Instead, they launched Project Glasswing with 12 major tech partners.
Anthropic just shipped Managed Agents — a fully hosted platform for running AI agents in the cloud. With OpenClaw sitting at 247K GitHub stars, the question is obvious: does the open-source darling still make sense?
OpenClaw is a free, open-source AI agent that connects LLMs to your actual software. It has 247k GitHub stars and a wild origin story involving trademark disputes and an OpenAI acquisition.
NemoClaw does not replace OpenClaw. It runs on top of it. Here is when you need one, the other, or both, and what the actual trade-offs are.
NVIDIA released NemoClaw, an open-source security stack that wraps OpenClaw AI agents in kernel-level sandboxing, local inference, and PII stripping. Here is what it actually does.
A straightforward walkthrough of getting NemoClaw running on Linux. One command installs it, a wizard configures it, and you get a sandboxed AI agent with PII stripping in about ten minutes.
Getting OpenClaw running takes about five minutes. Here is the full setup, from Node.js to your first working agent, plus the security settings you should change immediately.
OpenClaw has had three CVEs in 2026, a coordinated malware campaign on ClawHub, and over 135,000 exposed instances. Here is what went wrong and what you should do about it.
AI generated sites all look the same. The copy, the layout, the animations. Here's what gives them away and how to make yours feel human again.
Claude Code is changing how developers write software. Here's what it is, how it works, and why it's becoming part of every serious dev's toolkit.
Skills are reusable instruction sets that give Claude Code domain expertise. How they work, where to find them, and which repos are worth starring.
Turn Claude Code into a design-aware developer with these UI/UX skills. Color palettes, accessibility, responsive layouts, all in one place.
Build better web apps faster with Claude Code skills for Next.js, TypeScript, databases, authentication, and deployment.
React and React Native skills that enforce best practices, catch performance issues, and generate clean component code.
Whether you need a web app, mobile app, or AI-powered automation — let's talk about how I can help.
Get in touch