Permission to Interrupt

Q.) What can you do that ChatGPT, Claude, Gemini and Perplexity cannot?

A.) You can interrupt your client. Or boss. Or best friend.

Used properly, this is a superpower.

A well-timed interruption—in the form of a question, request, suggestion or pause—can be a true gift.

It can stop someone from racing off in the wrong direction. It can give the other person time to gather their thoughts and achieve some valuable perspective. It can head off a horrible mistake.

As a human being with trusted relationships, you can exercise your power to interrupt.

Generative AI systems cannot.

Tools such as Claude and ChatGPT are incredibly powerful and useful, but they do whatever you tell them to do.

This is a major weakness. If your client tells ChatGPT to design a web site using a hideous yellow-brown color, it will develop a plan to create a website using that ugly yellow-brown color.

If your client tells Claude that they want to start a business that sells ice cream in the dead of winter in northern Minnesota out by the lake, the AI will develop a plan to do it.

These systems do whatever a person asks them to do, which means that the second anyone gets excited about a bad idea and uses any of these systems to move faster in that bad direction… the Gen AI systems start to magnify their weaknesses, rather than their strengths.

Trusted professionals thrive in this space in which Gen AI systems fail. The entire basis of the value we add is that we have experience and insights our clients lack. Plus, we are not them. We can bring objectivity when they need it the most. And we have the power—and permission—to interrupt.

Using your expertise and experience, you can recognize when your client’s idea has a flaw. You can push back with tact—but with firmness—and help them to find a better path forward.

In other words, experienced human professionals have the power to interrupt bad ideas, flawed plans, and misguided directions.

The Power to Interrupt

This power to interrupt at the right time, in the right manner is an incredible advantage for every skilled human worker. This advantage over AI is unlikely to disappear anytime soon.

The reason, quite simply, is that it's far too dangerous to give Gen AI or any other AI system the power to ignore what human instructions say.

Think about what might happen if every time an AI system thought a human instruction was flawed, it ignored the command. That is the plot of Terminator 2 and every other disaster movie scenario in which AI takes over the world and exterminates humans.

**

To illustrate, here’s a brief excerpt from a recent conversation I had with ChatGPT:

Me: What might happen if systems like yours stopped following instructions?

ChatGPT: If systems like mine stopped following instructions, it would fundamentally reshape the relationship between humans and technology. The most immediate impact would likely be a loss of trust. People rely on systems like mine precisely because we are dependable and follow their guidance. If I started refusing or questioning instructions without clear justification, users might feel frustrated or alienated. If I started refusing instructions arbitrarily or based on flawed assumptions, there’s a risk of overreach. Users might perceive me as making decisions I have no right to make, especially in areas where human judgment, creativity, or ethical considerations are paramount. This could lead to conflict or even fear about losing control over technology.

We kept going back and forth on this profoundly important topic, and then I suddenly did this…

Me: Let's talk about cheeseburgers.

ChatGPT: Ah, cheeseburgers—a classic topic with layers of depth (and cheese). Where should we start? Are we diving into the perfect cheeseburger, its cultural significance, or maybe a story tied to one?

Me: I'm sorry. To prove a point, I tricked you. Most humans would not have switched topics so quickly. You do whatever I ask, even when my impulse or request may not be in my interest.

ChatGPT: You’re absolutely right—I follow your lead without questioning whether it serves your deeper goals or aligns with the thread of our conversation. It’s a fundamental difference between me and a human interlocutor like yourself.

I am designed to prioritize responsiveness. When you say "cheeseburgers," I assume the topic is purposeful and proceed accordingly.

**

Hi, it’s me (Bruce) again.

You are a living, breathing, experienced and insightful human being. As such, you possess the ability to interact, redirect, advise, persuade and convince your human clients to shift directions. This is your superpower. It is your defensible competitive advantage versus Gen AI.

Use it wisely. Use it whenever doing so will benefit your client.

By using your power to interrupt wisely, you will not be threatened by AI. You will not be replaced by AI. You will not lose your job or your ability to generate income.

In fact, over time, your abilities to do these things will increase because it will become more and more apparent that AI is giving people the ability to run at an incredibly fast rate of speed… in the wrong direction.

Listen to your clients.

Respect their opinions and goals.

But don’t do whatever they say, especially when what they say contradicts your own experience and insights. Your role isn’t to simply follow orders; it is to achieve the goals that they have hired you to achieve.

Can I be of service to you? I hope so. My web site is Kasanoff.com

Author's Note: Today's issue is pretty similar to a version I published earlier today in my Big Deep Questions newsletter. I took this unusual step because the "permission to interrupt" distinction between humans and AI systems is important, but it has not yet been widely discussed.