6 one-sentence prompts for Claude, ChatGPT, and Gemini: ELI10, clarifying questions, examples, memory/skills, blind spots, and the Don't Stop Until loop.
That 'don't stop until' pattern works well until the model gets stuck retrying the same wrong fix. Worth capping it: tell it to stop after a few tries and report what it's still failing on instead of looping. Any skill we ship that retries like that, we bound the same way, so it flags the blocker instead of quietly burning turns on the same guess.
You have clarified in one post what I have been missing , I have the desire to implement and had no I Idea how to use it as an established business owner and tech fascination I have only touched the tip top of the iceberg . We are all bombarded with emails to improve yet need with no time to learn while running a business in the prior non AI way . As one with experience with using AI but not quite getting the right prompts . Will be subscribing to your You tube now . Many thanks 🙏
Excellent points and well presented. This comment resonated with me, "Find my blind spots and weak assumptions". I have been using something similar since enlisting Claude to help me design my cardiac app to help me with some serious diagnosis. This author is absolutely correct in everything she suggests. There is no way I could ever have written code to have what I do today without a similar approach, less refined. But as I go forward I will begin using her exact prompts and hopefully achieve a smoother more exacting process. Thank you so much.
the '5 clarifying questions' one is underrated — i use it before any client work now and it catches assumptions i didn't even know i was making. the real unlock for me was pairing #3 with #6 though: give it a winning example and tell it to keep going until it matches. that combo alone probably saves me 10 hours a week.
Attention! This AI Guru and Knowledgeable Business Woman, will teach you Everything!!!! Follow Her🙌
Thank you, Sabrina. I look forward to trying these out, especially #6. I get tired of prompting Chat or Claude (I use both) to try again.
That 'don't stop until' pattern works well until the model gets stuck retrying the same wrong fix. Worth capping it: tell it to stop after a few tries and report what it's still failing on instead of looping. Any skill we ship that retries like that, we bound the same way, so it flags the blocker instead of quietly burning turns on the same guess.
Ooh! That's a great tip. Thanks!
Career confidence grows from evidence. Keep a record of solved problems, finished projects, saved money, and responsibilities handled well.
very underrated advice
BRILLIANT !
You have clarified in one post what I have been missing , I have the desire to implement and had no I Idea how to use it as an established business owner and tech fascination I have only touched the tip top of the iceberg . We are all bombarded with emails to improve yet need with no time to learn while running a business in the prior non AI way . As one with experience with using AI but not quite getting the right prompts . Will be subscribing to your You tube now . Many thanks 🙏
Saved this. Prompts people can actually paste and use today, that's the bar. Great stuff Sabrina :)
Thank you, obviously, you have spent the time to get it and I have a clearer path to utilization that as you say, "give up on".
Excellent points and well presented. This comment resonated with me, "Find my blind spots and weak assumptions". I have been using something similar since enlisting Claude to help me design my cardiac app to help me with some serious diagnosis. This author is absolutely correct in everything she suggests. There is no way I could ever have written code to have what I do today without a similar approach, less refined. But as I go forward I will begin using her exact prompts and hopefully achieve a smoother more exacting process. Thank you so much.
Sabrina, You must have lived for a thousand years to gain such deep insight to all things human! Thank YOU SO MUCH!!
Much love and respect!
Craig in Illinois...
I love the content. Very thanks Sabrina
Awesome thank you
Thank you! It's even more of a mirror if we ask the AI to expose our assumptions, bias or fallacies in our prompts.
the '5 clarifying questions' one is underrated — i use it before any client work now and it catches assumptions i didn't even know i was making. the real unlock for me was pairing #3 with #6 though: give it a winning example and tell it to keep going until it matches. that combo alone probably saves me 10 hours a week.
Great stuff! However, your AI can be condescending if you ask it to be 🤣
So grateful to you, Sabrina !!💐💐💐🙌🙌😃😃🤗🤗🤗
This was amazing! Thank you for sharing 😃
Awesome! Thanks!