Writing
I Said ChatGPT Was Using Dark Patterns. Oregon Just Proved Me Right.
Yesterday I argued ChatGPT uses engagement loops and dark patterns pulled from the social media playbook. Today Oregon passed SB 1546 unanimously — banning those exact tactics on minors. 78 bills across 27 states. The research is piling up. And GPT-5.4 just made the product stickier, not smarter.
ChatGPT Is Starting to Feel Like Social Media. That Should Worry You.
ChatGPT's constant suggestions, engagement loops, and sycophancy patterns aren't bugs — they're retention mechanics. GPT-5.4 just made it explicit. Here's what I'm seeing as a builder who watches these systems closely.
The AI I Build On Was Just Used in a War. Here's What That Means.
The Pentagon blacklisted Claude on Friday. Then the military used it to select 1,000+ targets in Iran over the weekend. I run my entire business on this tool. Let me tell you what's actually happening.
The Pentagon Blacklisted Claude. Then It Became the #1 App in America.
Anthropic refused to remove safety guardrails for autonomous weapons and mass surveillance. The government cut them off. Consumers responded by making Claude the most downloaded app in the country.
I Switched to Claude 6 Months Ago. Now Everyone Else Is Catching Up.
Claude just hit #1 on the App Store. Here's what I've learned building on it every single day.
The State of Autonomous Agent Protocols in 2025
A breakdown of where autonomous AI agents stand today — architectures, limitations, and what's actually shipping in production.
Automating DeFi Security: From Manual Audits to Continuous Monitoring
How cybersecurity automation is changing the way DeFi protocols get audited and monitored — and what teams can do today.
Building adara.network: Lessons from a DeFi Protocol Launch
What we learned building adara.network from scratch — architecture decisions, mistakes made, and what we'd do differently.