The latest note · September 17, 2026
Your Agent Believes Everything It Reads
Boris Cherny at Anthropic posted that prompt injection is largely solved on the newest models. Anthropic's own research page says progress, not solved. Both are right, and the reason why is the most useful thing to understand before you hand an AI agent your email, your CRM, and a browser.
Applied AI13 min read

23 notes since April 2026
The archive
Signal Notes is where I think out loud. Each piece breaks down a single marketing or AI topic, drawing from live client work, hands-on testing, or new research I'm currently digging into. Every note either hands you something to put to work or something I found too interesting to sit on. Usually both.
Next note · In the works





















