Inc.
AI can sharpen your thinking, but only if you don’t let it do all the thinking for you.
A Chatbot Can Give You a Polished Answer to the Wrong Question. These 3 Checks Prevent It
Author: Minda Zetlin
AI can make you smarter, if you use it right. It can provide research and analysis to augment your own understanding and help you make better decisions. But what’s the best way to blend AI’s power with the human judgment and experience only you bring?
Use what some experts call a “Think Sandwich,” advises leadership expert and author Erica Dhawan. In piece for CNBC.com, she offers five ways to be more memorable in the age of AI. All five are worth checking out, but I particularly like the idea of the sandwich. It goes like this: “Think, AI, Think.” Before you prompt AI to give you its analysis, begin by thinking about the problem you want to solve on your own. Then get AI’s expert input. Follow it up with your own analysis and judgment to arrive at a final decision.
For the first Think in the Think Sandwich, Dhawan suggests asking yourself three questions. They all make a lot of sense to me. Try them out, or adapt them for your own use, next time you’re about to turn to AI to analyze a situation, solve a problem, or help you draft an important communication.
1. What do I believe about this?
Starting with this question “forces you to develop your own point of view before AI gives you one,” Dhawan writes. You may not know what to think about the issue, and that’s OK. But even so, it’s worth taking a few moments to check your own gut feeling about the question, or better yet, spend a few minutes writing down whatever thoughts come to mind.
Unless you’ve specifically prompted it otherwise, AI will usually give you the prevailing or most commonly held view of whatever the topic is. That is, of course, a valuable point of view for you to consider. But by taking the time to think about the question before you ask AI for a response, you may come up with a viewpoint of your own that has nothing to do with the wisdom of the crowd. That can help you stand out, especially if everyone else is just relying on an AI answer.
2. What problem am I actually solving?
Not long ago, I was in a virtual session about using AI to be more productive. As part of the event, the workshop leader had an AI tool do a quick analysis of my work. It came back with a detailed analysis of how I was missing the obvious opportunity to use AI to help come up with ideas for columns. But I wasn’t looking for more column ideas. I was looking for ways to use my time more efficiently by handing some tasks off to AI.
It was a classic case of AI presenting a beautiful solution to a problem—but not the right problem. Asking yourself this before you prompt AI to help you “keeps you from getting a brilliant answer to the wrong question,” Dhawan writes.
3. What would a bad outcome look like?
Some AI tools have a definite positivity bias. They’re like a friend with a sunny disposition who always looks on the bright side and never considers the problems that could derail you or your project. So compensate for that bias by questioning the future before you ask AI for its take on the problem.
To do this most effectively, consider doing a brief premortem, a technique developed by cognitive psychologist Gary Klein. It requires you (or you and your team, or your advisors) to assume a future in which the project you’re about to start or the product you’re about to launch has failed. Now ask why it failed—what is the most likely reason? Asking this question “makes you consider the risks and consequences that a polished AI response might overlook,” Dhawan writes.
There’s no question that AI can help you make better decisions and perhaps improve your outcomes. By asking these questions and creating a Think Sandwich, you take full advantage of those capabilities. And you also bring the benefit of your own unique perspective and experiences to create a solution that’s stronger than either you or AI could have created alone.
Credits: Project Syndicate