Summary In today's episode, I explain why text prompts often fail to produce consistent AI images and how you can overcome this limitation. Here's what this means for you. You can achieve much higher visual stability across multiple generations by changing your approach to prompting. You'll also learn these concepts: why language remains too imprecise…
I listened to no podcasts today on my drive. You’re probably like, “And?” Normally when I do long distance drives, I catch up on podcasts. Today, I didn’t. Today, I listened to the NotebookLM summaries of papers I needed to catch up on, advanced AI papers from NeurIPS that I didn’t understand well. Over the…
Summary In today's episode, I explain how to craft effective prompts for complex tasks by using task decomposition. Here's what this means for you. You gain the ability to move beyond generic AI responses to produce unique, repeatable, and high-quality results. You'll also learn these concepts: how task decomposition prevents model confusion, why specific cognitive…
Context engineering is data management. That’s anticlimactic, isn’t it? You hear everyone from Satya Nadella (Microsoft CEO) to AI bros hyping “context engineering” as the next big thing. It’s data management. If prompt engineering is building great recipes, context engineering is building great ingredients. Even the most decorated Michelin-starred chef can’t make rotted, spoiled food…
Summary In today's episode, I explain how to troubleshoot specific errors or loops in AI responses. Here's what this means for you. You can better diagnose why a model fails and use more effective prompting to get accurate results. You'll also learn these concepts: the distinction between reasoning and non-reasoning models, the advantages of using…
Prompt engineering is hot again. Not that it ever went away, but there was a hot minute there where every AI expert was saying that prompting was dead and no one would ever need to learn prompting. Well, that was before agents. And I don’t just mean the big, heavy, super technical ones. ChatGPT Agent…
Summary In today's episode, I show you how to stop generative AI from using American spelling and culturalisms. Here's what this means for you. You can ensure your AI output maintains your specific regional linguistic identity. You'll also learn these concepts: why reasoning models improve accuracy, how positive prompting outperforms negative commands, and how checklist…
_Almost Timely News: 🗞️ How to Audit Your Marketing Strategy with AI (2026-01-11) :: View in Browser_ The Big Plug 🚨 Take this one question survey to tell me and the Trust Insights team what webinar/training content you’d like us to make this year! Content Authenticity Statement 100% of this week’s newsletter was generated by…
Ever had that friend who takes everything literally? You love them, you know them, and you know there are certain things either you just shouldn’t ask them, or you have to be very clear and specific how you ask them. That’s AI. Many of our personal language biases, AI takes literally. And because it’s built…
Summary In today's episode, I explain how to prevent AI models from losing context or memory during long conversations. Here's what this means for you. You can maintain highly accurate and relevant AI interactions by managing how you provide information to the model. You'll also learn these concepts: the limitations of the AI context window,…
