Thoughts on the agentic web, AI security, and the future of internet traffic.
34 tactics from 19 research papers. How to format, cite, frame, and structure content so AI trusts it — with data, models tested, and dates for every tactic.
New research shows LLMs ignore your descriptions, trust government over social media, and flip judgments based on framing. Here's what 19 studies say about convincing AI.
Formatting swings LLM accuracy by 76 points. Fake citations fool GPT-4 more than humans. Here's what persuades AI and why it's nothing like persuading people.
The internet now has two audiences: humans and AI agents. Switch is building the infrastructure to design two webs on one site. Here is our 3-phase roadmap, what is live today, and where we are headed.
The internet just got a billion new users: AI agents. Here is the framework (and 30-day plan) to make your site readable, logical, and actionable for them.