ChatGPT doesn’t quite get complexity (at least not according to Einstein!)
Update: Before reading on, a warning. ChatGPT creates references out of thin air and presents them with great confidence as factual. You cannot trust a reference provided by the bot, you HAVE to double check. I found non-existant papers attributed to real authors; real papers attributed to the wrong authors … and many other combinations. So be warned!
You are probably thinking: not another ChatGPT post! It seems to be on everyone’s mind nowadays. I have realised though that regardless of whether I believe it will prove beneficial, detrimental, or even dangerous to humanity, I can’t afford NOT to be curious about it. So I’ve played around with it, trying different prompts. Some of these interactions were useful (it’s not bad at creating marketing copy!), others not so much.
One example … I thought I’d see if it can help me with a challenge I’ve set myself: to get better at explaining complexity in the most accessible ways possible. I am doing this because I believe that the wisdom of complexity needs to become more widely available; at the risk of sounding trite, it needs to be democratised. It is pretty challenging to explain what complexity is without using the complicated language and jargon without dumbing it down and sacrificing theoretical rigour. This is one reason why I often revert to using analogies like contrasting zoos and jungles — which also has clear limitations.
So what happened with ChatGPT:
I took inspiration from this quote attributed to Albert Einstein …
“If you can’t explain it to a six-year-old, then you don’t understand it yourself”.
And asked ChatGPT: How would you explain the difference between complicated and complex to a 6-year-old?
Here’s how the conversation went …
ChatGPT: “Let’s think about a toy. A toy can be complicated if it has a lot of small pieces that you have to put together, like a model car. It might take a long time to build and it might be hard to figure out how all the pieces fit together. But, once you build it, it’s not hard to play with, it’s just hard to build.
On the other hand, a toy can be complex if it does a lot of different things, and it’s not always easy to understand how it works. Like a toy robot that can walk, talk, and dance all at the same time. It’s hard to understand how all those things work together, it’s complex. And it’s also hard to play with it because it’s not always clear what buttons to press or how to make it do what you want.”
This was a bit underwhelming!
ME: That didn’t make the difference particularly clear to me …
ChatGPT: “Think of it this way: a complicated toy is one that has a lot of pieces and instructions to put it together, while a complex toy is one that has many interconnecting parts that work together to do something interesting or fun.
For example, a lego set can be considered complicated because it has many small pieces and instructions to follow, but once you put it together, it becomes a complex toy because all the pieces fit together to create a unique and interesting structure.”
Still not great. So I changed tack.
ME: How would you explain “complex adaptive”?
ChatGPT: “Complex adaptive means that there are many pieces that work together to make something happen, and those pieces can change and adjust based on how they interact with each other. Imagine a big ant colony where all the ants work together to build their home and find food. Each ant does its own job, but how they do their jobs can change based on what other ants are doing. This makes the whole colony work together in a complex way.”
Me: Better, but I’m not sure a 6-year old will understand that.
ChatGPT: “Okay, let’s try a simpler example. Imagine playing with a set of blocks. You and your friend each have a bunch of blocks, and you’re building something together. As you build, you notice that the blocks you add affect what your friend is adding, and vice versa. You both keep adjusting and changing your building based on what the other person is doing, until you end up with a big, complex structure. That’s like complex adaptive: the blocks are adapting and changing based on how they interact with each other, to create a bigger and more interesting structure together.”
Me: Could you use family as an example?
ChatGPT: “Sure! Imagine a family as a group of people who live together and interact with each other. Each person in the family has their own thoughts, feelings, and behaviors, but they also affect and are affected by the other people in the family. Just like the blocks in the previous example, each person in the family is adapting and changing based on their interactions with the others, leading to a complex and dynamic family system. This is similar to what we mean by complex adaptive systems.”
Me: Please illustrate that with a short narrative appropriate for a 6-year-old.
ChatGPT: “Sure! Imagine a family of three: a mom, a dad, and a little boy. The mom is cooking dinner, the dad is reading a book, and the little boy is playing with his toys.
While the mom is cooking, she notices that the little boy is getting hungry, so she starts making more food. The dad notices that the little boy is getting restless, so he puts down his book and starts playing with him. The little boy sees that his dad is playing with him, so he starts feeling happier and more relaxed.
All three of them are changing and adapting based on their interactions with each other, just like building blocks or ants in a colony. This makes their family work together in a complex way, just like we mean by complex adaptive systems.”
This was just terrible! I decided to take a break from this.
I wanted to see how it did with other target audiences.
Me: How would you explain the difference between complex and complicated to Economics students?
“Complex and complicated are often used interchangeably, but there is a subtle difference between the two.
Complicated refers to a system that has many parts, features, or details, and is often difficult to understand, analyze, or predict due to its many interdependent elements. For example, a tax code can be considered complicated because it contains many rules and regulations that can be difficult for someone to understand and follow.
Complex, on the other hand, refers to a system that is composed of many interconnected and interdependent parts, where the relationships and interactions between these parts give rise to new and emergent behaviors. For example, an economy can be considered complex because it is composed of many interconnected actors (e.g. consumers, businesses, governments) who interact in complex ways to produce emergent behaviors like market trends, economic cycles, and financial crises.
In summary, a complicated system is simply difficult to understand, while a complex system is composed of many interdependent and interconnected parts that give rise to new and emergent behaviors.”
Me: and CEO’s?
“For CEO’s of large organizations, the distinction between complex and complicated can have significant implications for decision making and organizational strategy.
Complicated refers to a system that has many parts, features, or details, and is often difficult to understand, analyze, or predict due to its many interdependent elements. For example, a supply chain that involves many suppliers, intermediaries, and logistics can be considered complicated, as it can be difficult to understand the impact of changes or disruptions in one part of the chain on the overall system.
Complex, on the other hand, refers to a system that is composed of many interconnected and interdependent parts, where the relationships and interactions between these parts give rise to new and emergent behaviors. For example, a rapidly changing market with many competitors and new technologies can be considered complex, as it requires a deep understanding of the interconnections and interdependencies between the various actors in the market to make effective decisions.
In summary, a complicated system is simply difficult to understand, while a complex system is composed of many interdependent and interconnected parts that give rise to new and emergent behaviors. As a CEO, it’s important to understand the difference between complicated and complex systems in order to make effective decisions and drive organizational success.”
Better, although I don’t agree with it’s examples e.g. supply chains are not merely complicated …
I then returned to the 6-year-old but with a different approach. I told ChatCPT to assume the persona of a children’s book author like Dr Seuss. It then came up with this terrible rhyme, although I do like some of the concepts.
“Once upon a time,
in a world of twists and turns,
There lived a tiny being,
with a curious mind, yearning to learn.
It noticed that some things in life, were like a maze,
With twists and turns, that would leave one in a daze.
And then it found, that some things were like a neat machine,
With gears that turned, and always ran as clean.
These were called, the neat and orderly,
With rules and plans,
they were quite predictable and steady.
The tiny being loved to explore, and found beauty in both,
For each one was unique, and held its own truth.
It learned that the maze, was full of surprises,
And held secrets, and adventures, that would arise.
And it saw that the machine,
was simple and straightforward,
But sometimes lacked the spark,
of life’s own artward.
So the tiny being came to understand,
That both the maze and the machine,
have a place in this land.
For each one is special,
in its own way,
And that’s what makes life,
an exciting play.
So embrace the twists and turns,
and the simplicity too,
For they both hold wonders, waiting just for you.
For life is a blend of both, the orderly and the maze,
And it’s up to you, to appreciate and embrace.
My conclusion: ChatGPT is useful as a sounding board or to come up with initial ideas. It doesn’t come close to human creativity (in my opinion), and it’s NOT reliably accurate … someone who doesnt’ know the difference between complex and complicated would come away from this exchange confused or misinformed. More worrying, it’s propensity to create false references means that its responses has to be double checked. So instead of saving time, you end up spending more time searching for purely made-up articles and authors.
While ChatGPT probably won’t steal our jobs anytime soon, someone comfortable with the technology might. So it is critical that we find ways to use it (and others like it) generatively and evolve with it. It also makes it even more important to keep challenging ourselves to find better ways to connect with our audiences, keep our content fresh and build trust so that readers don’t fall for AI generated content so easily.