BY STEPHEN SHAPIRO
INNOVATION CONSULTANT AND SPEAKER
Like so many others, I have been playing around with ChatGPT, the chatbot A.I. offering from OpenAI. It does a reasonable job of writing articles (or even books), answering questions, and analyzing information.
But can it innovate? I decided to put it to the test.
Before starting, it's important to agree on a definition of innovation. For this article, I've decided to focus solely on problem-solving and problem reframing, with an emphasis on the latter. True innovation is often the result of asking better questions that lead to better solutions.
Given this, I asked ChatGPT to reframe a problem I frequently use as an example:
I work for an airport. People are complaining about how long they have to wait at baggage claim for their checked bags. Is there a better way to look at this problem?
Those who have heard me tell this story know that one possible solution is, instead of speeding up the bags, you can slow down the passengers. Make them walk further. This reduces wait time at baggage claim.
Did the A.I. "think" of this as an option? Here are its responses:
There are a few ways you can approach this problem. Here are a few ideas:
Analyze the baggage handling process: Look at how bags are being sorted and delivered to the baggage claim area. Are there bottlenecks or inefficiencies in the process that could be causing delays?
Monitor baggage claim wait times: Keep track of how long it takes for bags to arrive at the baggage claim area, and use this data to identify patterns or trends. This can help you identify problem areas and come up with solutions.
Communicate with travelers: Make sure travelers know what to expect when it comes to baggage claim wait times. Provide information about average wait times, and let travelers know if there are any delays or issues that may affect their experience.
Consider implementing technology solutions: There are several technology solutions that can help improve the baggage handling process, such as RFID bag tags and automated sorting systems. These can help reduce wait times and improve the overall experience for travelers.
Work with airlines and other partners: Collaborate with airlines and other partners to find ways to improve the baggage handling process. This could involve coordinating baggage handling procedures or implementing new technologies.
Not bad. Logical. Nothing overly insightful, although No. 3 was clever in that it was not specifically about the speed of the bags. But the other responses were far from innovative.
I then asked it explicitly to reframe the problem again. The answers were nearly identical. I then explicitly asked it to reframe the problem of "reducing wait time" and that again didn't help.
My last attempt was to try and "teach" it to think. I took a chapter from one of my books where I tell the entire baggage claim story. In it, I suggest that there are different ways to tackle the problem, including slowing down the passengers, reducing the wait time, improving the wait experience, or distracting passengers while they wait. But the starting point was, "How can we speed up the bags?"
Could the A.I. learn how to reframe using my story as an example? I asked it:
Given what you now know, please reframe the original problem "How can we speed up the bags?" 7 different ways.
Sure! Here are 7 different ways to reframe the original problem of "How can we speed up the bags?":
How can we reduce wait times at baggage claim?
How can we improve the overall experience of waiting for luggage at the airport?
How can we ensure that bags and passengers arrive at baggage claim at the same time?
How can we minimize the time it takes for luggage to be transported from the plane to the baggage carousel?
How can we streamline the process of handling and transporting luggage within the airport?
How can we optimize the speed and efficiency of the baggage handling system?
How can we reduce the amount of time it takes for travelers to receive their luggage after arriving at their destination?
Nos. 1 and 2 are straight out of my story. It still didn't pick up on slowing down the passengers, although No. 3 hints at it if you read between the lines. Nos. 4 through 7 are just the original "speed up the bags" problem said differently.
I did more experiments, but in the end, the system wasn't fully up to the task of reframing. It helped, but it didn't generate anything extremely thought-provoking.
Of course, a system like this is only as good as the questions that are asked of it. And maybe there were better questions I could have asked or better stories I could have used to train it.
In a world with greater adoption of technology, the World Economic Forum listed innovation and complex problem-solving as some of the top skills needed in business for 2025. This feels spot on. Technology is of course incredibly useful. However, in a world where A.I. becomes widespread, many human skills are still critical for business success.