Sunday, November 15, 2026

Before you start, check with AI! (Part Two)

Let’s now go over that three-stage approach and see how it works. 

To get started, let’s start a process in which we keep asking increasingly specific questions. We would then check important facts against reliable sources and, when appropriate, qualified professionals, helping us make our decision with a much better understanding of the situation. 

The second stage would give us more context and make AI more useful. For example, ask: "What should I know about replacing a fence?" and explaining: "I’m replacing a six-foot backyard fence on a sloping property. Here are the contractor's specifications, the materials proposed, the soil conditions, and the competing bids. What am I overlooking, and what should I challenge?" 

An additional benefit of AI is that it would help us overcome our blind spots or our own biases, and that might actually be one of AI's greatest contributions. We all have blind spots and preconceived notions, including those of us with considerable experience. 

Sometimes we are so accustomed to doing things a certain way that we don't even consider alternatives. AI can introduce an element of constructive opposition, by digging deeper and responding to these questions: 

  • What am I overlooking? 
  • What could go wrong? 
  • What assumptions am I making that could be incorrect? 
  • Is there a less expensive or more effective alternative? 
  • If we were fully responsible for this project, what would concern you most? 
  • Uncover the strongest argument against our preferred solution. 

This last point is a particularly key question that forces us to look beyond our initial convictions. Of course, AI can also be confidently wrong, so its answers need verification, particularly for medical, legal, financial, engineering, and safety-related matters. To that end, I would add one more dimension that could make AI our preliminary thinking partner. 

Let’s explore it as our next blog concludes that series...

No comments: