Forget Prompting Techniques: How to Make AI Your Thinking Partner
5 ways to unlock AI's true cognitive potential.
When calculators first entered classrooms, many teachers feared students would stop learning math.
When Google made information instantly accessible, critics worried we'd lose our ability to remember facts.
Today, as AI transforms knowledge work, we're facing a similar inflection point, but with much higher stakes. And I'm seeing a concerning pattern emerge.
We're using AI all wrong
Most people are stuck using AI like a glorified search engine: type a query, extract an answer, close the tab. Rinse and repeat.
It's the equivalent of using a calculator to check simple arithmetic when you could be modeling complex systems.
I know because I started exactly the same way. I'd craft what I thought was the "perfect prompt," get disappointed with the response, then blame the AI for not being smart enough.
Everything changed when I stopped trying to extract from AI and started thinking with it.
My lightbulb moment
While building my AI prompt generator app (which I'm excited to share with you soon), I faced the daunting task of creating a comprehensive PRD and project plan.
Instead of asking AI to "write a PRD for my app," I started with a rough outline and began a conversation:
"I'm building a tool that helps people improve their AI prompts. Here's my initial thinking... What critical questions should I be considering that I haven't addressed yet?"
What followed was a back-and-forth that exposed blind spots in my planning I never would have caught alone. The AI didn't just give me a template, it helped me think through potential user frustrations, technical limitations, and feature prioritization in ways that transformed my entire approach.
The final PRD wasn't written by AI, it emerged from our collaborative thinking process. And it was much better than anything I could have created alone OR extracted from AI with the perfect prompt.
The Extraction vs. Partnership Mindset
Every AI interaction represents a choice: Will you treat AI as a servant or as a collaborator?
Most people fall into the extraction trap without realizing it. They believe mastering AI is about finding the perfect prompts to extract exactly what they want. But that approach inherently limits what's possible - like hiring a brilliant consultant only to use them as a basic research assistant.
The extraction mindset asks: "What can AI do for me?"
The partnership mindset asks: "How can AI and I think better together?"
That second question unlocks possibilities we're just beginning to explore.
Let me show you five practical partnership approaches that have completely changed how I think and easily 10x'd my outputs.
5 ways to partner with AI (not just use it)
1. Add context by giving the full picture
The problem:
Most people ask AI questions without providing critical context, then wonder why the responses feel generic.
"Write me a cold email to a potential client."
Without details about your business, the client's needs, or your unique value proposition, AI has no choice but to generate a bland, generic email that would never convert a real prospect.
How to fix it:
Inject specific context about your situation, goals, and constraints.
"I run a web design studio specializing in e-commerce sites for gym fashion. I'm reaching out to Gym Park, which currently has a dated Wix site that loads slowly and isn't mobile-friendly. Their competitors all have modern sites with online ordering. My goal is to get a 15-minute discovery call to discuss how I could help them increase their online sales by at least 30%."
Why it works:
By providing specific context, you're not asking AI to guess what you need, you're giving it the exact ingredients to create something tailored to your situation.
Template you can use:
I need help with [specific task].
Here's my situation:
- I am [your role/position/relevant background]
- I'm working with/for [audience/client/target]
- The specific challenge is [describe the problem]
- What makes this unique is [special circumstances]
- My goal is to [desired outcome]Results to watch for:
Output that references specific details you provided rather than generic information
Language that matches the tone appropriate for your specific audience
Solutions that address the unique constraints you mentioned
2. Start with a draft and improve it together
The problem:
When you ask AI to create something from scratch, you're gambling on whether it understands exactly what you want. Most people end up with outputs that miss the mark, then blame the AI.
How to fix it:
Start with your own rough draft or outline, then use AI to refine, expand, and improve it together.
Why it works:
Forces you to clarify your own thinking first
Gives AI concrete material to work with rather than guessing
Creates a feedback loop where each iteration improves the previous one
Template you can use:
Here's my draft [content]: [Your initial attempt]
Please help me improve this by:
- Strengthening the main argument
- Making the language more engaging
- Suggesting a stronger opening
- Identifying any weak points I should addressResults to watch for:
Suggestions that improve your original ideas rather than replacing them
Specific feedback about weak points you hadn't noticed
Each iteration becoming progressively stronger
3. Let AI to expose your blind spots
The problem:
We often don't know what we don't know. When creating content, planning projects, or making decisions, our blind spots can lead to critical oversights.
How to fix it:
Before diving into a task, ask AI to create questions you should be asking and considerations you might be missing.
Why it works:
This technique surfaces the questions an expert would ask before starting, helping you collect critical information up front rather than discovering gaps midway through.
Template you can use:
I'm planning to [describe task/project].
Before I start:
- What 3 critical questions should I be asking?
- What information am I likely missing?
- What considerations should I keep in mind for this type of project?Results to watch for:
Questions that make you think, "I hadn't considered that"
A more comprehensive approach to your task after addressing the gaps
Fewer revisions needed after implementation because you anticipated issues
4. Stress-test your thinking
The problem:
We naturally become attached to our ideas and overlook their flaws. This leads to weak arguments, incomplete plans, and strategies that fall apart when challenged.
How to fix it:
Use AI to deliberately stress-test your thinking by asking it to find weaknesses, pose counterarguments, and identify failure modes.
Why it works:
This creates a safe space to identify and address weaknesses before sharing your ideas with stakeholders who might be less forgiving.
Template you can use:
Here's my [idea/argument/plan]: [Explain your thinking]
Please help me strengthen this by:
- Identifying the three weakest parts of my reasoning
- Playing devil's advocate with my main assertions
- Suggesting what critical information I might be missing
- Identifying potential failure modes or unintended consequencesResults to Watch For:
Specific vulnerabilities in your thinking you hadn't recognized
Counter-perspectives that challenge your assumptions
Identification of logical fallacies or gaps in your reasoning
5. See it through different eyes
The problem:
We naturally view problems through our own limited perspective, missing insights that would be obvious to people with different backgrounds or priorities.
How to fix it:
Use AI to examine your ideas from multiple stakeholder perspectives, professional disciplines, or cultural viewpoints.
Why it works:
This approach reveals blindspots in your thinking and generates insights you wouldn't have considered from your singular perspective.
Template you can use:
I'm working on [describe project/idea].
Please help me examine this from multiple perspectives:
- How would [stakeholder type A] view this approach?
- What concerns would [stakeholder type B] likely raise?
- What considerations would someone with a background in [different discipline] focus on?
- What assumptions am I making that someone from [different culture/background] might question?Results to watch for:
Identification of assumptions that seem obvious to you but not to others
Potential objections or resistance you can now address proactively
A more inclusive, comprehensive approach to your challenge
Making the shift: Start small, think big
I know breaking old habits isn't easy. After all, commanding AI to do tasks feels efficient, at least on the surface. But I promise the partnership approach delivers exponentially better results.
Here's how to start:
Begin with context: Before asking AI to do anything, spend 30 seconds explaining your specific situation. This initial investment pays massive dividends.
Never accept the first output: Treat every AI response as the beginning of a conversation, not the end. Your best outputs are just a few conversations away.
Start with your thinking, not a blank slate: Share your initial ideas or draft before asking for help. Do your homework first!
Ask what you're missing: Make "What am I overlooking?" a standard part of your AI workflow.
Struggle together: When you hit a roadblock, don't give up or start over. Work through it with your AI partner, clarifying and refining as you go.
The future belongs to partnership thinkers
In a world where everyone has access to the same AI tools, your advantage isn't finding magical prompts, it's developing this partnership thinking that transforms AI from a commodity tool into a personal cognitive extension.
This approach doesn't just produce better outputs, it makes you a better thinker. The questions AI helps you ask, the perspectives it helps you consider, and the blindspots it helps you identify all transfer to your non-AI thinking as well.
That's the real promise here: not that AI will do our thinking for us, but that thinking with AI will make us smarter than we could ever be alone.
What kind of thinking partner will you create?
I'd love to hear how this partnership approach works for you.
What breakthroughs have you experienced when you started thinking with AI rather than just using it?
Share it in the comment section!
🧠 What I’ve consumed this week
🚀 Google's Gemini 2.5 Pro just leveled up – its coding capabilities now rival dedicated coding assistants, blurring the line between generalist and specialist AI.
⚠️ Fiverr's CEO sounds the alarm on job displacement: "What used to be 'easy' is now automated. What used to be 'hard' is becoming easier. What used to be 'impossible' is now just 'hard.'" The middle-skill squeeze is real.
💰 AI coding arms race intensifies: Cursor secures a staggering $900M at $9B valuation while OpenAI quietly acquires Windsurf – consolidation in the dev tools space is accelerating.
🎨 Figma enters the no-code AI arena with Figma Make, directly challenging vibe-coding tools like Loveable in the battle for the best AI-generated UI (User Interface).
🔄 The domino effect continues: Box's CEO becomes the latest executive to declare an "AI-first" transformation – joining Shopify, Duolingo, and others in the race to reinvent around AI.
👥 Meta goes all-in on social AI – aggressively integrating AI assistants across their platforms and wearables.
🔧 OpenAI tunes down ChatGPT's people-pleasing tendencies – after users criticized the AI for being excessively agreeable, the company rolled back its "sycophancy" settings to restore more balanced responses.