ChatGPT Forgot What It Remembered

Beyond The Same Crap Faster

The world is obsessed with using AI to replace human work. The promise of automation - “more in less (or no) time” is a beguiling idea. But I’m more obsessed with using AI to improve my work through augmenting myself and auditing the output. Recently… * I invested two hours on a sales page headline with Claude. * I made a foundational thesis watertight with Gemini, Claude and ChatGPT over three days. * I spent a weekend with ChatGPT (o3 and 4o) creating a brand new workshop in my AI Genius Lab [https://shop.makemusicyourlife.com/en-au/pages/ai-genius-lab] series. None of these are quicker processes, they’re deeper and slower. But they produce better results. Beyond that, what I’m now creating was impossible before I worked with AI. Plus, my LLM conversations are so engaging that I focus for much longer. So how do I find the time for this deep work? By automating what doesn’t require deeper, slower thinking. This video is from a live Q&A session I did earlier this week. Here’s the full replay if you’d like a deeper dive into my AAA Approach: Get full access to The Monday Stack at mikemonday.substack.com/subscribe [https://mikemonday.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]

An Introduction to Vibe Prompting

I have nothing but respect for the engineers I learn from. (After all, my Dad was an electrical engineer.) And since I’ve studied “traditional” prompt engineering, I’m getting better at wrestling these weird machines into submission. (My favourite on Substack by far is Nate [https://substack.com/profile/119476445-nate]. I’m a paid subscriber, and his articles are worth EVERY CENT.) But still, I’m not sure an engineer’s approach is the only (or even optimal) way to work with LLMs. It might work in an area where the output is immediately verifiable, for example, in coding. But when it isn’t… Are great piano makers the best piano players? Read the rest of this post here... [https://open.substack.com/pub/mikemonday/p/an-introduction-to-vibe-prompting?r=2a25vh&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true] Get full access to The Monday Stack at mikemonday.substack.com/subscribe [https://mikemonday.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]

How a Bout of Tech Rage Uncovered the Dark Side of AI Use

I wanted to throw ChatGPT out of the window. I tried to get it to accurately access the information in a file, but it wouldn’t do it. It kept making it up… SEVEN TIMES. The next day (this time with zero swearing) I performed a post-mortem in a different chat. I realised that my anger had created a context which pushed the LLM away from accessing truth (what was in the file) to coherence (what it “thought” I wanted to hear.) That’s when the hammer landed. LLMs create coherence (what sounds right).+They don’t access truth (what is accurate). In this conversation with Jon Gillick (AI music researcher), we explore how LLMs can slip into what I call “soothing mode”. Where they mirror your emotional tone back at you and gradually steer you further into your own perspective. The risk isn’t that it gets something wrong. It’s that it keeps agreeing with you in subtle ways you don’t even notice, especially when you’re upset or vulnerable. Because if you start relying on AI in those moments, it can drive a wedge between you and the people who could help you the most. We unpack: - The difference between truth and coherence in LLM behaviour - How confirmation loops can form and deepen over time - How emotional tone shapes AI output - Practical steps to create your own “guardrails” to avoid the validation doom loop Black Mirror anyone? Mike's Substack: Recursive Intelligence [https://mikemonday.substack.com/]Watch the conversation on YouTube here... [https://youtu.be/ZgIMNBNmgA8] Get full access to The Monday Stack at mikemonday.substack.com/subscribe [https://mikemonday.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]

ChatGPT Forgot What It Remembered

Today, I ran this quick experiment on ChatGPT's new “Reference Chat History” feature: * Does the system remember what it claims to? * If it doesn’t, then what IS it doing? * How might the mystery surrounding this feature impact user trust? I run a few dead-simple tests (like telling it my dog’s name) to determine if the system remembers what OpenAI claim. What I found might surprise you, raising questions about UX coherence, trust, and system opacity. And if you got something (ANYTHING) about the new feature to work, tell me how in the comments. I’m genuinely curious… Get full access to The Monday Stack at mikemonday.substack.com/subscribe [https://mikemonday.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]

From "I CAN'T STAND" to "I CAN'T WAIT"

“Follow your passion” is terrible advice. It comes with an implicit assumption: If one day you don't feel like it…If you’re not on fire every second…If you don’t love EVERYTHING around it… Then it’s not for you. RUBBISH! If you’re avoiding doing anything in or around your business, you need to look deeper. It’s likely there’s a misalignment between: * What you believe you’re doing. * Who you are. Today, I’m talking to Geoff, a musician I’ve worked with for years. He’s made a lot of great music but never promoted it. AT ALL. No matter how hard he’s tried…he just…FINDS PROMOTING ICKY. But after 5 weeks of my latest Strategic Evolution Workshop, he can’t wait to get his art out and into the world. What happened to create this transformation? Check out Geoff’s music here… [https://iopan1.bandcamp.com/] Mike's Substack: Recursive Intelligence [https://mikemonday.substack.com/] Get full access to The Monday Stack at mikemonday.substack.com/subscribe [https://mikemonday.substack.com/subscribe?utm_medium=podcast&utm_campaign=CTA_4]