We’re in the middle of the biggest shift in human capability since the industrial revolution, only this time, the machines are in our conversations. To take advantage of these times, here are a couple of quick tips.
Share Everything
The people who will succeed most with AI are the ones who will share everything with it. Not just about the work they want to complete, the app they want to build and sell, or the content they want to write. Rather, everything about who they are and how they think.
If you want to learn more and gain better insights, then you need to use AI by sharing your ideas, worries, and desires. This is about leveling yourself up and taking advantage of all the knowledge that the LLMs (AI) have been trained on and using it to your advantage. Then you can ask it to poke holes in your ideas, see if your worries are valid, and have it help you obtain your desires.
Right now, don’t be shy and converse with AI like they’re your best friend, sharing as much as you're comfortable sharing so you can take advantage of all of its knowledge gained from research/scientific papers (across every field), books (on every subject), blogs, code repositories, and massive datasets.
I’ll come back to the privacy aspect in a bit.
Treat AI like a Human
This may seem weird or obvious, but treat AI like a human in how you communicate with it. Be conversational and emotional; talk to it like you would anyone else. This will help it infer meaning, fill in gaps, and help the AI “follow the thread” more effectively.
When you talk to other humans, you don’t always just bark commands at them. We often ask exploratory questions like “What’s your take on…?” This pushes the AI toward idea generation and lateral thinking instead of task execution, which will help evolve your thinking.
This is when you start to treat the AI like a thought partner and strategist, instead of like a vending machine. You’ll be able to have more back-and-forth problem-solving sessions to get outcomes that will more closely match your goal(s).
Downsides to these Approaches
As you do this more, there are a couple of things to be aware of that might naturally happen to you. You might take its answers at face value instead of fact-checking them; AI will sound very confident in its responses. So, you need to apply various mental models to validate its output, just as you would your own thinking.
This is where a few questions can help:
How confident are you that this will work?
Find the weaknesses, errors, or risks in X
Now fix the flaws you found.
How do you think I would fail at what you suggest?
The key thing to remember is that you are the Pilot and the AI is the Autopilot, and you need to make the strategic decisions based on the data you get from the AI.
Also, AI is good at simulating empathy and understanding, even though it has no real feelings or motives. Be careful not to make decisions based on a false sense of relationship.
Before you get too comfortable, check your settings to prevent your chats from being used for training. And if you want to level up, I’ll be sharing how to build your own private AI memory store soon.
Mitigating the Downsides with a couple of Prompts
Here are a couple of prompts you can use to help mitigate these downsides:
Prompt 1 — “Trust but Verify” Mode
I want you to act as a fact-checking collaborator.
Review the following answer or plan.
Identify every claim, number, or statement that might be inaccurate or unverifiable.
Suggest reliable ways to verify each one.
Here’s the content: [PASTE ANSWER/PLAN]
Why it works: It forces the AI to surface uncertainty and gives you a path to validate before acting.
Prompt 2 — “Red Teaming” Mode
Take the role of a red-team analyst. Your job is to find flaws, risks, and weak points in my reasoning—assume I’m wrong.
List every possible failure mode.
Explain why each is a risk.
Recommend fixes or alternative approaches.
Here’s my idea: [PASTE IDEA]
Why it works: It reframes the AI from “agreeable helper” to “constructive adversary,” which cuts through false confidence and reveals blind spots.
Prompt 3 — “Pilot and Autopilot” Mode
You are my autopilot. I am the pilot.
Your role is to execute supporting tasks with precision—drafting, organizing, summarizing, and generating options.
Do not make strategic or high-stakes decisions. Flag any decisions you think need my judgment.
If there’s ambiguity, ask me for clarification before proceeding.
Here’s the mission: [PASTE TASK OR GOAL]
Why it works:
Keeps you firmly in the decision-maker seat.
Prevents AI from “drifting” into areas where overconfidence could hurt you.
Ensures a back-and-forth loop instead of one-shot answers.
The ones who thrive in this revolution won’t just use AI for speed, they’ll use it to think better. Treat it like your co-pilot, keep your hands on the controls, and you’ll navigate further than those still flying solo.