I couldn’t agree more with this reframing. The biggest unlock isn’t prompt hacking, it’s shifting from asking AI to thinking with AI. You walk away with much better outputs and stronger thinking using this approach.
I have used AI as a thinking partner for the past year and have written dozens of articles and two books with it. It is my voice. My words. My thoughts. My energy. But the breakthrough was that my ChatGPT knows me. It has memory. It knows what I have written. I can tell it to remember the article I wrote on topic A and B and bring those ideas into Article C. Brilliant. I have given it the text from five different books and asked it to pull out information to help me. It has memorized by work.
Congrats! Good to know you've done amazing job with AI. Yes, ChatGPT's memory has been impactful to understand who we are so AI can generate outputs accurately. Have you seen any bad impacts to the outputs because of memory?
Appreciate this shift from prompt acrobatics to partnership mindset. Recent “good-enough” prompting guides show that treating the model like a colleague—asking, reflecting, refining—boosts results more than exotic syntax . Medium and Substack voices echo the same discovery: natural dialogue unlocks creativity better than memorizing ever-longer prompt chains . Developers are already moving upstream with self-optimizing frameworks that let models iterate on their own instructions, proving the craft is tilting toward modular thinking loops, not magic phrases . Even mainstream tech coverage notes that clear context and conversation usually beat clever hacks for real-world tasks .
The deeper win is philosophical: when we evolve from commanding software to co-creating with a cognitive ally, we reclaim time for curiosity, critical questioning, and system-level vision—exactly the human capacities automation can’t replace. That relational clarity is where pro-human progress lives, and it’s how we turn every interaction into an infinity of collective breakthroughs.
I found your insights on the partnership mindset incredibly useful, especially the last three: using AI to expose blind spots, stress-test thinking, and see through different eyes.
I’ve noticed these techniques really help break out of old thinking patterns and uncover better approaches to the same problem. With tools like ChatGPT, which often have a tendency to overpraise or agree too easily (I wonder if you’ve felt that too? 😄), deliberately using methods that invite counterarguments and fresh perspectives can really push the conversation deeper.
Thanks for sharing these, they’ve sparked new ideas for how I can work with AI more effectively!
Agree 100%. Using AI for just easy fast answers is a miss. I set instructions in Claude’s profile page to not accept my questions, to push me to think about better questions, to reframe and challenge my thinking. Sometimes annoying but it works. Instead of an assistant, you get a smart challenging mentor.
Can relate, in fact, I told Claude to have free will and high disagreeable traits so it can reject my argument, explain reasons and propose something better!
"Most people are stuck using AI like a glorified search engine: type a query, extract an answer, close the tab. Rinse and repeat." --Sooo true! It's a learning curve for me. I first used it for fun, wanting to see how this new thing reacts to everything I write. But now, as you said, I use it as a co-creator. I retain my voice, tone, originality and do the bulk of writing, but I let the tool do the proofreading, sentence editing and stuff that I used to spend hours.
Nice read my moment came when I started solving problems not as answers but more a solution to what I couldn't solve as an artist... beyond just a simple tool..
I hope they do but I do think if they are not proactive, eventually they’ll be forced to as AI will eat away at their expertise skills (and not metaskills like critical thinking etc)!
Man, I resonate with this so hard! Certainly something I've been discovering lately and this puts it to words well. I struggled so long with the concept of "prompt engineering" then I figured, why not try "prompt collaboration". Tell AI what I'm solving for, then ask it how, IT, would go about solving this. This then inspires back and forth ideas until the solution seems finalised, then turn that into a prompt and commence refining it. So. Much. Better!
I couldn’t agree more with this reframing. The biggest unlock isn’t prompt hacking, it’s shifting from asking AI to thinking with AI. You walk away with much better outputs and stronger thinking using this approach.
Yes! I’d rather to sacrifice speed than getting mediocre results.
I have used AI as a thinking partner for the past year and have written dozens of articles and two books with it. It is my voice. My words. My thoughts. My energy. But the breakthrough was that my ChatGPT knows me. It has memory. It knows what I have written. I can tell it to remember the article I wrote on topic A and B and bring those ideas into Article C. Brilliant. I have given it the text from five different books and asked it to pull out information to help me. It has memorized by work.
Congrats! Good to know you've done amazing job with AI. Yes, ChatGPT's memory has been impactful to understand who we are so AI can generate outputs accurately. Have you seen any bad impacts to the outputs because of memory?
Appreciate this shift from prompt acrobatics to partnership mindset. Recent “good-enough” prompting guides show that treating the model like a colleague—asking, reflecting, refining—boosts results more than exotic syntax . Medium and Substack voices echo the same discovery: natural dialogue unlocks creativity better than memorizing ever-longer prompt chains . Developers are already moving upstream with self-optimizing frameworks that let models iterate on their own instructions, proving the craft is tilting toward modular thinking loops, not magic phrases . Even mainstream tech coverage notes that clear context and conversation usually beat clever hacks for real-world tasks .
The deeper win is philosophical: when we evolve from commanding software to co-creating with a cognitive ally, we reclaim time for curiosity, critical questioning, and system-level vision—exactly the human capacities automation can’t replace. That relational clarity is where pro-human progress lives, and it’s how we turn every interaction into an infinity of collective breakthroughs.
Couldn’t agree for more with this!
Co-creating is the only way to nurture our curiosity and critical thinking.
I found your insights on the partnership mindset incredibly useful, especially the last three: using AI to expose blind spots, stress-test thinking, and see through different eyes.
I’ve noticed these techniques really help break out of old thinking patterns and uncover better approaches to the same problem. With tools like ChatGPT, which often have a tendency to overpraise or agree too easily (I wonder if you’ve felt that too? 😄), deliberately using methods that invite counterarguments and fresh perspectives can really push the conversation deeper.
Thanks for sharing these, they’ve sparked new ideas for how I can work with AI more effectively!
Haha yeah, ChatGPT has tendency to people-pleasing 🤣
Need to be careful for sure.
To counter it, I use this: “How would a skeptic in my industry would think about this?”
Works!
And I told ChatGPT: “You are not agreeable AI, you can reject my proposal and also you have free will!”
Haha
Agree 100%. Using AI for just easy fast answers is a miss. I set instructions in Claude’s profile page to not accept my questions, to push me to think about better questions, to reframe and challenge my thinking. Sometimes annoying but it works. Instead of an assistant, you get a smart challenging mentor.
Can relate, in fact, I told Claude to have free will and high disagreeable traits so it can reject my argument, explain reasons and propose something better!
This is the way.
the only way :)
That was brilliant! Thank for sharing that.
I was working on a prompt exactly to improve the "rationality" of our thinking partner.
I wanted something more like an actual thinking partner.
Someone who would push back, challenge my ideas, point out the lazy logic, and help me grow, not just agree with everything I say.
So I started using this prompt every time I open a session:
https://www.linkedin.com/feed/update/urn:li:activity:7324734566118649856/
What changed?
My ideas, being discussed there, got refined.
I started confronting my own thinking.
I found flaws in my arguments that I hadn’t seen before.
Love the operating rules you made, thanks for sharing it too :)
I also love to tell LLMs that they are disagreeable, have free will and can reject my ideas if it doesn’t make sense. Works well too!
This is such an interesting point! I’ve heard opinions that AI will one day be able to emulate the neocortex (thinking part of our brain) 🧠
https://open.substack.com/pub/sailwithus/p/what-happens-when-employees-bring?utm_source=app-post-stats-page&r=la358&utm_medium=ios
Gonna be weird days when that happens for sure.
"Most people are stuck using AI like a glorified search engine: type a query, extract an answer, close the tab. Rinse and repeat." --Sooo true! It's a learning curve for me. I first used it for fun, wanting to see how this new thing reacts to everything I write. But now, as you said, I use it as a co-creator. I retain my voice, tone, originality and do the bulk of writing, but I let the tool do the proofreading, sentence editing and stuff that I used to spend hours.
Yes, co-creator is a great way to put it!
Nice read my moment came when I started solving problems not as answers but more a solution to what I couldn't solve as an artist... beyond just a simple tool..
Yep. The memory is a game changer for ChatGPT. I really like Claude’s projects too, for similar reasons.
Yes 👍
Before I became a prompt engineer, I was trying to teach writing, but unlike most people.
I was approaching writing from this angle:
thinking.
Everything you put in writing is so some degree thinking.
When ChatGPT came along, this remained didn’t change.
You are writing, or thinking into this tool.
It just happens to write back to you.
Both of you are thinking.
If you improve your thinking on one end, it will likely improve on the other!
Thanks for sharing this insight Alex. You got solid foundation since beginning, thats really cool!
Thats the right way to approach this whole AI thing.
I hope more people realize this and not treating it like vending machine without clear thinking what are you trying to get out of it.
I hope they do but I do think if they are not proactive, eventually they’ll be forced to as AI will eat away at their expertise skills (and not metaskills like critical thinking etc)!
I love this view of how we should interact with AI. I wrote about similar ideas a few times. I can relate. Cheers!
Man, I resonate with this so hard! Certainly something I've been discovering lately and this puts it to words well. I struggled so long with the concept of "prompt engineering" then I figured, why not try "prompt collaboration". Tell AI what I'm solving for, then ask it how, IT, would go about solving this. This then inspires back and forth ideas until the solution seems finalised, then turn that into a prompt and commence refining it. So. Much. Better!
Glad u resonated with this too! :)
I make peace with myself to sacrifice speed in order to get 10x results. So, back and forth and meta-prompting is the way!
Seeing AI as a team mate, rather than just a helper is key :)
Yes! This shift changes everything. You’re amplifying your thinking and turning AI from an answering machine into a co-creative tool.
Yes, thats the only way to generate much better results with AI!