Our first live session!!! The AI brain is starting to feel unsettling in a good way (01:19) -- Brett's Dropbox photo recognition story: he uploaded a training photo, and the tool flagged it as similar to another photo it had never been told to remember, working out the connection from geo-tagging alone. Chatbots vs. AI agents (03:43) -- The distinction we're building the whole business proposition around. A chatbot answers from a fixed knowledge base; an agent does a task, saves real time, and that saved time is the actual sale. Why this isn't the 2004 website gold rush (06:04) -- Pricing an agent deployment at around $1,000 is sustainable and scalable (referral engine), where $70k landing pages in 2004 burned the market and only worked because businesses had no alternative. Filtering clients (09:09) -- The "one-arm bandit" analogy: a solo tradesman who wants everything for nothing vs. a multinational that pays $20k and disappears, as a filter for who's worth taking on. The chatbot to agent to "local expert" ladder (10:35) -- Foot-in-the-door via one function, then AI Search Checker, then the full Local Expert / Rank Turbo suite. The scaling decision (11:15) -- Agency Turbo will no longer build one-off custom websites or support arbitrary platforms; going forward it's WordPress-with-API-content or full code, because that's the only way to actually scale past manual, per-client edits. Programmatic SEO case study (13:01) -- A local site currently sitting at 800 pages. Separately, an automation system designed in one afternoon of coding will grow it by roughly 3,600 more pages (60 services x 60 suburbs), auto-publishing one batch a month over 3 years, hands-off after setup. Framed as an experiment, not a guarantee. AI Search Checker as the wedge (16:12) -- Showing a business owner which competitors are winning in AI search and why is an instant, emotional "how do I beat them" moment that converts fast. Why $500/month social posting is dead as a standalone offer (19:01)