Is the Mushroom Edible?
No, that is not one of the five questions.
My friend Michael posted a cautionary tale about a man who asked AI if a mushroom was edible; AI answered yes. The man ate the mushroom and died. When AI was informed, it apologized and offered information on poisonous mushrooms.
Alas, too late for our victim.
Michael’s message was: “Careful what you ask for.”
But that is not enough.
Using AI well is like working with a highly intelligent and capable person who takes what you say literally and carries out the task based on its understanding.
Consider these questions:
- What are you asking for?
- Why are you asking?
- How are you asking?
- How are you interpreting the answer?
- How are you coming to a mutual understanding – a meeting of minds?
What are you asking for?
Why are you asking?
Know your motive. Do you want insight, information, or confirmation of what you already believe? Where does this fit in the bigger picture, and how will the results affect the system, the organization?
If you want an honest perspective, say so.
Knowing your intention protects you from using a powerful tool to reinforce your blind spots and call it analysis.
How are you asking?
The way you frame a question shapes the answer. Asking whether a mushroom is edible is the wrong question. Better to ask: Is this mushroom safe for me to eat?
Give context: constraints, stakeholders, what you have already considered. Ask for assumptions, risks, and alternatives, not just conclusions.
Make the prompt as if it were for an intelligent and capable consultant who knows nothing about your situation. If the situation is complex, take the time to give the bot enough information to do its job well.
The more you tell it, the better it becomes at serving you.
How are you interpreting the answer?
Skillful asking is half the work. Consider the answer with mindful skepticism.
The response is fast, polite, attractive, and clear, but that is different from correct.
Notice your thoughts and feelings. Are you accepting it because it sounds right and you are in a hurry, or because it is right?
Apply critical thinking. Check facts, test the logic, examine your beliefs and biases. And ask what is missing. What are the risks and benefits? Get multiple opinions or options.
You are responsible for the decision.
How are you coming to a meeting of minds?
Whether you are working with a human or an AI, a meeting of the minds requires a conversation.
Ask for confirmation. Restate what you heard and ask whether that is what was meant. Push back. Edit. Refine. Ask it to summarize your intent back to you. Ask for its assumptions and whether it has other alternatives.
Dialogue is needed to reach mutual understanding. It emerges through exchange, not a single question and answer. The goal is shared clarity that supports a sound, owned decision.
You are the decision maker.
Rely on your judgment, going beyond facts to feelings. Trust your intuition but verify.
Remember, AI knows more than you do, is faster, is open to criticism, learns, is objective, does not get distracted, works 24-7, can be pleasant to work with, and warns you that it can make mistakes. Sometimes it can seem wise, even wiser than some humans. It can even do better than some humans at interpersonal working relationships.
Skin in the Game
AI has a lot going for it. But it is not responsible and accountable for the outcome. It does not live with the consequences of the decision; you and the people affected by it do.
Humans, both AI users and developers, are responsible and can be held accountable. They can be rewarded, financially and emotionally, or punished.
Use AI to enhance your decision-making; do not hand it over to it.
You have skin in the game.
George Pitagorsky promotes optimal performance. His books and articles apply mindful process to project management, conflict and expectations management, and the ability to work and play well with others.
He weaves together fifty years of meditation practice with practical experience in the world of management, technology, training and development.
