We all live through various hype cycles, personally and professionally. (Hype cycles look a lot like the Satir Change model, the image in this post.) Right now, in 2024, “generative” AI is having its moment.
But AI is not generating anything, certainly not new insights. Instead, the engines have ingested (without paying creators) lots of words and images. These models are terrific at pattern-matching. That’s why AI can work for diagnostic medicine, especially if the model offers percentage probabilities. It’s also why AI appears to be so useful for text and images.
Pattern-matching is the easy stuff—regurgitating what we might have understood if we had more time to read and see “everything.” (I’m not suggesting that this pattern matching does not take time to do right, especially for diagnostic purposes.)
When Mark planned our Northern Lights trip last March, he used AI to find places that fit our needs. AI was a tool to save us time—the easy part.
I have no objection to saving time on the easy parts. That’s the whole point of tools. But generating insights and curating our humanity? So far, that’s a human endeavor.
People Generate Insights (a Hard Part)
So many people (managers mostly) think that they can replace developers or testers with AI. However, that’s not how product development works. Technical work is not about the typing. Instead, people take unrelated ideas and combine them to create innovative products.
That’s a hard part.
It’s also hard because people, us fallible humans, need to understand what other people have done before. Otherwise, we turn into a cargo cult. Cargo cults carry out practices without understand of the reasons behind those practices.
Right now, only humans can integrate those unrelated ideas and create products. (To be honest, most of the AI “successes” I’ve seen are all “meh.” The writing, images, and products don’t inspire me. But, that’s me.)
How do humans generate insights? By working together, in a culture or environment that focuses on how people can work together. Tools can help, but the culture or environment work is really hard. That’s because it requires humanity to be effective.
Our Humanity is a Hard Part
Regardless of my role, I’ve been most effective when I work with people. Inside the organization, I helped build bridges of understanding between other humans. As a consultant, I collaborate with my clients on everything, from deciding what problems to solve and how I will deliver those “solutions.”
We can work as humans when we have a culture or environment that supports collaboration. That’s quite hard. (I write about that on my other site because it’s so, so challenging.)
Tools can help, especially pattern-matching tools. Just imagine if an AI tool said to a manager, “I realize you want to start more projects. But high Work in Progress leads to lower throughput. You won’t get these projects when you want them.”
Or, imagine an AI tool that said, “Stack-ranking has been disproven. Here are 20 references that explain why it doesn’t work. Instead, consider these three alternatives to manage salary costs and performance.”
I would love a tool that did that! (I might not have to write more management books, though, which would disappoint me. Heck, I’d write them anyway.)
We need to recognize what’s hard and what’s easy. Tools can confuse our thinking.
Clarify the Hard Parts and the Easy Parts
Tools always automate the easy parts. While I’m not using AI because of the copyright issues, I might in the future. Count me in for using tools to make my work and life easier.
But the human parts of our work are never easy. They’re so, so hard. That’s because we are not deterministic. Organizations can never be like well-oiled machines and our lives are full of randomness. Remember the old Yiddish saying, “Man plans and God laughs.”
Work and life, creation and delivery—those are the hard parts because we are interdependent humans. AI might be a great tool for the easy parts. But the hype cycle will never make it easier to be human. The more AI pervades our world, the more humanity we need to work and live together.
Humans create culture. It’s a delightfully challenging hard part because we need our adaptability. Not just to live through this hype cycle, but to integrate the new things when they and we are ready.
Let’s not confuse the hard parts with the easy parts.
I use AI because it produces predictably daft results. I have a feature on my blog where the world’s worst advice columnist gives terrible advice. The hard part is creating a funny prompt for the AI to fail to understand and utterly miss the subtext. Training it to focus on irrelevant and daft things was the easy bit. This only works because AI is nearly useless at giving life advice and okay about following unreasonable directions. I do the inventive bits and the AI does the typing stuff no sane human would say.
That’s hysterical. Thank you so much for this. (I’m laughing out loud!)