Mastering AI Collaboration: 8 Key Techniques
For those of you following Terminator GMAT’s Instagram, you might know I’ve been intensely focused on developing AI study tools. I’ve been pushing AI to its message limit daily, which unfortunately led to tendonitis in my finger. I just got a 3800 NTD steroid shot at the hospital today – ouch!
But why am I so passionate about this? Because in my interactions with AI, I’ve found endless enjoyment and rediscovered the essence of efficient learning.
Today, I want to share the mindset and methods you need to effectively leverage AI.
Mindset 1: Reclaim Your Inner Scholar — Play Dumb, Ask Relentlessly, and Challenge Authority
不確定自己現在幾分?
12 題免費測出分數區間 →
When interacting with AI, I put myself back in the role of a student. I found myself instinctively asking question after question, patiently reviewing its detailed answers, and then digging deeper into anything that seemed questionable. I’d push it to explain concepts in different ways, even asking it to break them down for a middle schooler.
In reality, I already knew how to solve the problem and where its logic might have gone wrong. But it’s precisely through this attitude of playing dumb, asking relentlessly, dissecting every detail, and not fearing authority that I could train the AI to develop correct thinking patterns.
Everyone, please remember: play dumb and ask relentlessly, listen patiently, dissect every detail, and most importantly, “do not fear authority.”
Mindset 2: Master AI’s “Converge” and “Diverge” Frameworks
Once you’ve adopted the right mindset, you can start applying practical methods. I’ve organized the ways I use AI in life and learning into two main categories and eight key techniques:
- Category 1: Converge
- Techniques: “Collect,” “Organize,” “Analyze,” “Summarize”
- Category 2: Diverge
- Techniques: “Brainstorm,” “Create,” “Imitate,” “Develop”
The essence of mastering AI lies in constantly switching between “converging” and “diverging.”
Practical Exercise: A “Converge” and “Diverge” Cycle
- Converge: First, I fed the sample score report and my independent analysis results to the AI, asking it to “organize” my key analysis points and “compare” the data, telling me what aspects I might have missed.
- Diverge: Regarding the missed aspects, I asked the AI to “brainstorm” how I should assess them and provide additional suggestions.
- Converge: I incorporated the AI’s suggestions into my original analysis process, distilling it into a new analysis guidebook.
- Diverge: Using the spirit of this guidebook, I “created” new prompts for the AI to “imitate” me and analyze a new score report.
- Converge: I once again “compared” the AI’s analysis results with my own independent output, and then fine-tuned my analysis guidebook.
- Diverge: Finally, I had the AI help “develop” this refined guidebook into an automated VBA program to assist me in future analyses more efficiently.
Throughout this process, you must constantly maintain the spirit of “asking deeply.” Never give up on communication just because you think something is “impossible” or “not good enough.” If you feel the result is poor, tell the AI specifically what’s wrong so it can correct it. If you don’t know how to communicate, you can also ask it: “What steps do I need to go through to achieve your goal?” Let it guide you step by step.
Conclusion: Your Flexible Thinking is the Key to Unleashing AI’s Power
Do you see it? Effective communication with AI is essentially a continuous process of converging and diverging. Your role in this is indispensable; your flexible thinking is the key to unlocking AI’s potential as a strategic advisor.
I’m thrilled that after more than a decade in this field, I remain a passionate educator, eager to embrace new things. Thinking about this makes me feel truly fortunate to be able to work with such enthusiasm and dedication.

