Hey, I'm Antonio, based in Helsinki. I run Google Ads campaigns and do SEO work, plus trying to get two businesses off the ground (agency + travel company). The travel one is basically me being stubborn about a market I know well. AI tools have become my unfair advantage for research, ad testing, content workflows, basically anything repetitive. Though half the time I'm still figuring out if I'm using them right. Excited to learn from what everyone's actually doing here, not just what sounds good on paper. I'll try to join Ryan's meetups when the timezone works out... Thanks for having me. Antonio

After the last online meetup, I was trying to track down where I’d read about Gemini integrating with NotebookLM. This looks to be the piece I had in mind:https://www.androidcentral.com/apps-software/googles-gemini-now-integrates-seamlessly-with-notebooklm-for-improved-project-management I’ve not tried it myself yet, so I can’t speak to real-world use cases or where it saves time. If you test it, let us know what you used it for and what worked (or didn’t).

Hey guys - First of all I apologize for anyone who was planning on attending the New Member Onboarding Call today. I had to cancel due to a chaotic week of a family death on Friday and unexpectedly putting our dog down last night 💔 I appreciate all of you for understanding and 2026 will be a great year for this community! I have a lot of good things planned. I hope you all have a great Christmas/Holiday Season with your families ❤️

I was just sorting out some of my bookmarks/favourites and came across this link I had saved a while ago. It's obviously been updated for 5.2, and if you've not already seen it, it's worth a read. https://cookbook.openai.com/examples/gpt-5/gpt-5-2_prompting_guide. I've yet to look up guides for others like Gemini, Claude etc. What AI had to say about this guide 'This guide introduces GPT-5.2, OpenAI’s advanced model for enterprise and agent workflows. It improves accuracy, instruction-following, concise formatting, and structured reasoning over previous versions, making it ideal for complex tasks like coding, document analysis, finance, and multi-tool operations. The guide offers best practices for writing effective prompts: set clear output length and content limits, prevent scope drift, handle long documents by summarizing, address ambiguity by clarifying or listing interpretations, and always ground answers in provided or cited sources. It also covers techniques for context compaction, agent updates, tool usage, structured data extraction, prompt migration, and in-depth web research—optimizing GPT-5.2 for reliability, consistency, and high-quality results in production settings.' There is also the link to the Prompt Optimiser that's been around for a while too. AI summarised this about it 'The Prompt Optimizer referenced in the GPT-5.2 Prompting Guide is a tool built into the OpenAI Playground. It’s specifically designed to help users improve existing prompts and facilitate migration across OpenAI models, including from GPT-5 and earlier versions to GPT-5.2.' Hope it's helpful for those who have not yet come across it.

Whats up everyone, quick question. I am getting so much better with script writing using AI. One of the things that I have done is create multiple GPT's that focus on different things. For example, the first one makes the entire script - its purely creative and broad. I copy all the chapters onto a google doc once I have an entire script. Then re paste that script into another GPT that focuses on adding curiosity loops, etc.. all the things we would do to make it engaging content. It rewrites it with those things in mind. Then, finally.. I put the entire script through one last GPT that makes it sound more human. So it again, updates the script a little bit to add a human touch and remove parts that sound robotic. I am wondering if I am over engineering this. The workflow is being done by a staff member so it also is a great way to outsource the work and know that even a non english person V.A can create an amazing output since the GPT's are doing all the heavy lifting. Has anyone done this before? AS in, use multiple prompts or GPT"S for the same long form content to refine it and make it better? Just curious... I kinda just discovered this workflow as I was interacting with a GPT asking it questions about improving output.