Early in my studies about AI, I learn to get around drifts and hallucinations by replacing text prompts with prompts that are in the style of formal symbolic formats. The end results were more predictable, in that I was able to control the order in which the data was processed (It was another method used to control token usage)

One of the things that really intrigues me is how different industries are using AI. I am working very hard in the financial industry right now and they run into a lot of red tape. Whereas a lot of my course creators and small business owners are free to use whatever they want. So what industry are you in and what sort of rules and regulations do you have when it comes to using AI tools? ps- if any of you are in finance or wealth management, send me a message and I'll send you the new prompt guide I just made for this industry

I asked Copilot to write something for me and include a link to my LinkedIn profile. You would think it would get that right since it is a Microsoft product, but it used an old LinkedIn URL I changed about three years ago. So I corrected it by saying, “That is not the correct URL. This is the right one, and please remember it for the future.” Results "Memory updated" What I appreciate is that Copilot can learn from corrections like this. If it gets something wrong, just tell it the correct information and ask it to remember it going forward. Small adjustments like this make the experience much more accurate and personalized.

The Prompt Formula That Crushes! [Role] + [Task] + [Context] + [Format] + [Constraints] + [Example] Example: "You are an email copywriter for digital product Write a follow-up email for prospects who opened but didn't reply to our initial outreach. Context: They're CTOs at 50-200 person companies, evaluating dev tools. Format: - Subject line (under 50 chars) - Opening (one sentence referencing their open) - Value prop (what's in it for them) - Soft CTA (no pressure) Constraints: 100 words max, conversational tone, no corporate jargon. Example tone: 'Saw you opened my last email. No pressure, but I built something that helps CTOs cut deployment time by 40%. Worth a 15-min call?'" Did you know: - Prompts with questions in them get 2.3x better responses - Adding "explain your reasoning" improves output quality by ~40% - Shorter prompts ≠ worse results (sweet spot is 50-150 words) - Prompts that reference previous context work better in conversations What's your favorite prompt or prompt formula?