You might think there are no stupid questions, but stupid answers certainly exist, and ChatGPT excels at them. Shep Hyken discovered this when he was surprised to see ChatGPT's inability to handle a stinker.

 

Text: Shep Hyken

As I write this article, I am on the CX Malaysia summit, where I was the opening speaker, and where I listened to Dr. Shreekant Vijaykar of COPC Inc., one of the other speakers, talk about AI and how you can't always trust its answers.

Dr. Vijaykar said he had asked ChatGPT the following question:
I just bought a pair of shoes from Amazon, and they delivered the wrong order. The shoe for my right foot is for my left foot, and the shoe for my left foot is for my right foot. What should I do about this error?

 

Dr. Vijaykar subsequently stated that ChatGPT's response was to contact Amazon to inform them of the error, after which they would either refund the purchase price or exchange the shoes.

I didn't believe Dr. Vijaykar, so I asked ChatGPT the same question. To my surprise, this was the answer:

That's certainly a strange mix-up—and something Amazon needs to fix quickly. Here's what you need to do:

Start the return/replacement process: Go to Your Orders on Amazon. Find the shoe order and click "Return or Replace Items." Select the reason: "Wrong item shipped" or "Item defective or not as described." Amazon usually provides a free return label and a replacement option.

Describe the problem clearly: Explain in the comments exactly what happened, for example, "The right shoe is shaped for the left foot and vice versa. The pair I received doesn't fit properly." This clarity helps Amazon's quality control team bring the issue to the seller or manufacturer's attention.

Choose a replacement (if you still want the shoes): If you like the shoes and they're still available, request a replacement instead of a refund. Amazon often ships these before your return arrives.

 

I took this ridiculous response from ChatGPT one step further and wrote:
"Maybe there's a better answer. It's not really a mistake. ... Just put the shoes on the other foot and they'll fit just fine. What do you think?"

ChatGPT's response was:
"Ha! That's a funny way to look at it, but unfortunately, shoes don't work quite that way. Even if you can technically wear them on the other foot, they'll be uncomfortable and can disrupt your posture or cause pain after a while, especially if you wear them for extended periods."

Dr. Vijaykar's experiment revealed an important truth about AI: it makes mistakes. ChatGPT treated two regular shoes as if they were defective and provided a detailed explanation of Amazon's return process.

AI is powerful, but just like humans, it can make mistakes. This ridiculous exchange is why some customers are afraid to use AI for customer support. If you use AI tools like ChatGPT, Claude, or others, you need a way to verify answers. Apparently, common sense isn't something that can be programmed. At least, not yet!

 

(Ziptone/Shep Hyken)

This article was originally published on Hyken.com.

Customer Experience, Featured, Opinions