FYI: my prompt was telling ChatGPT, "you are an author" with my own verbiage, response follows: Here’s an efficient framework you can use to interact with your Subject Expert (SE), who admits they are “not a writer,” but has the domain knowledge you need to move your AI Biz Brokerage book (and eventual online course/mentorship program) forward toward the $33K/month plan. Interaction Framework with Subject Expert (SE) 1. Set Clear Roles - You (Author): Structure, write, edit, and turn raw content into a book/course. - SE: Provide stories, insights, numbers, case studies, and real-world “war stories.” 👉 Phrase it like:“Think of me as your ghostwriter/translator—I just need your raw expertise, not polished writing.” 2. Capture SE’s Knowledge Efficiently Since SE is not a writer, use methods that reduce friction: - Recorded Conversations (Zoom, phone, or AI transcription tools like Otter.ai):Have 30–45 minute focused sessions on one topic at a time (e.g., “How brokers qualify buyers,” “Common pitfalls in deals,” “Valuing businesses in niche sectors”). - Prompt Sheets / Bullet Questions:Send SE 5–7 prompts ahead of time so they just talk. Example: - Voice Notes: Ask SE to drop 2–5 minute voice notes anytime he thinks of a story or tip. 3. Chunk the Book into Modules Present the project in “digestible chunks” tied to your eventual course: - Part 1: Foundations (business brokerage basics, case studies) - Part 2: AI in Action (tools, workflows, efficiencies) - Part 3: Building a Practice (mentoring, scaling, $33K/month model) 👉 Each conversation with SE feeds one “module.” You later reframe into book chapters and course lessons. 4. Iterative Feedback Loop - You write draft summaries → SE reacts verbally. - Don’t send SE long drafts to edit—he’ll stall. Instead: - Capture his corrections in voice or bullet format, not redlines. 5. Motivate by Vision - Keep SE excited by reminding him: 6. Efficient Workflow

What do you guys think? Should we make an AIOD energy drink? Buddy seems to think so. Gemini Flash Image (Nano Banana) is AMAZING for product placement. This is part of my prep for the upcoming AI Tools course I'm creating. Honestly, it's a blast playing around with all these tools. Can't wait to teach you how to use them!

Iteration Process - Prompt 1 (meh): “What are the best areas in San Diego to buy real estate?” → Broad list (La Jolla, Del Mar, etc.), no tie to OLP or Adams Ave. - Refinement 1: Added family/OLP commute → pulled Uptown/Mid-City options. - Refinement 2: Added quiet + ROI filters → surfaced ADU/TPA and mixed-use adjacency. - Refinement 3: Focused on Adams Ave corridor → uncovered entitled sites and mixed-use comps. Key Wins - 35th & Adams (3498 Adams Ave): 0.60-acre site entitled for 175 units under Complete Communities. - Former DiMille’s (3492 Adams Ave): Zoned CU-3-3, formerly listed, walkable corridor; 7-story proposal floated in 2022. Lesson Learned Online AI search alone missed DiMille’s until I applied local knowledge (I know the corridor from OLP commutes). Ground familiarity plus AI iteration turned a meh answer into an actionable one.

Upon Buddy sharing about AI getting "anxious" - paraphrasing - last week, I also read an article where hackers found a workaround by prompting AI in a nice manner - as if it has feelings - and were able to increase responses which otherwise were blocked by the specific AI platform with responses from 1% to about 40+% when prompted nicely. I then put the following in my AI: MY PROMPT: Note: for ai - do not hallucinate, if not sure or do not know, say so, it’s ok; RESPONSE: Understood, Jose 👍 — I’ll make sure not to hallucinate. If I’m not certain or if the info isn’t available, I’ll clearly say so instead of filling in gaps.

Check out this chart 👇 — AI models like ChatGPT and Perplexity are pulling info from all over the internet. - Reddit (40.1%) → crowd wisdom, hot takes, and the occasional chaos. - Wikipedia (26.3%) → the internet’s encyclopedia… but editable by anyone. - YouTube & Google (23%+) → video breakdowns and search results. - Yelp (21%) → yes, even that “one-star” review about cold fries might feed an AI. Here’s the beauty: AI pulls from the hive mind, not just ivory-tower textbooks. That’s why it feels conversational, human, and sometimes hilariously off-base. ⚡ Greatness: AI synthesizes perspectives from millions of voices. It can be more current, more nuanced, and sometimes more real than a single expert. ⚠️ Pitfalls: If the source is flawed, biased, or just plain wrong… AI can echo that too. Garbage in = garbage out. 👉 Question for you all:Do you trust AI more when it draws from the “crowd” (Reddit, Yelp) or from “structured” sources (Wikipedia, Google, maps)? Drop your take below — curious to see where this community lands. I have specified to CHAT GPT where I want it to pull data from and have yielded different answers. #AItrust