The short answer is no. The long answer is that our intuition becomes better over time.
I have seen a lot of chatter and PDQs that the job is shifting to require deeper domain knowledge. I challenge the assumption that this makes us more accurate. Sometimes, being a domain expert inhibits our ability to step outside ourselves and put ourselves in our customers shoes. We are meant to be question askers who can quickly come up to speed on different domains. And we’re meant to validate our assumptions by putting our customer front and center. That is what really helps us get it right!
Does AI always get it right?
The same short answer is no. It is a helpful tool as long as you remember that transformers like ChatGPT (Chatty) are prediction engines. They are taking the data they have been trained on, or that you have provided, and providing a contextually appropriate answer. Which means it still have to be validated by putting our customers front and center.
What is the solution?
I predict that the Product Manager will remain a viable and valuable position. And those who can adapt to using AI will be the most successful.
An example.
I have been working on a portfolio project that imagines a new book tracking application, similar to a Goodreads or StoryGraph model. AI has been invaluable. I have been feeding Chatty product development methodologies, primarily Hypothesis-Driven Development and Lean Startup. We’ve also been conducting research together on the market, users, and product needs.
As I prepared to move into experiment design and experimentation, I had to make a critical resource decision. Namely, with no money to support the research and limited personal reach, how could I come up with a reasonable facsimile of the experiment results. Chatty was the answer. We both had the same input information and knew the goals. I asked Chatty to take that and come up with the psychographic profiles for my ideated personas. Once I had that, I conducted the interview as if I was that persona, imagining what they would say.
The Results
Even with the same inputs, our profiles/interview for the author segmentation were off and I had to go back to the persona assumptions. Further research also determined that the reader segment core dimensions were continuing to shift. After refinement of existing profiles that failed, tightening profiles that passed, and imaging the interview results of the new personas, Chatty and I have come up with a tight and prioritized list of personas and proven our persona hypotheses.
Check out my portfolio to see the detailed experiment design and experiment results for round 1 and round 2 testing.
My Life in Practice 
Leave a comment