For a few years now, I’ve tried to tell people (especially PMs) that the most important thing they should be trying to develop is product intuition. I’ve said the same things that I think many, many others have said
Good product intuition is as much (if not more) art than science
the best product decisions aren’t always directly explainable in the moment - though somehow people will find an explanation for them in retrospect. :-)
the only way to really know if you’re good at making product decisions (or really any decisions) is to look back every 6 months to a year, and see what decisions worked out, what you reversed, and what really paid off. If you’re mostly happy with them, you’re finally getting good at making them.
some people are actually naturally better at it then others..
… but you can definitely get better at “product intuition” or just “decision-making” in general - some combination of experience, thinking about how you make errors, and looking at lots of data and examples does always help you get much better fast, but I’ve found how, and how much, to vary a lot by each person.
Whenever, I explained this to people they mostly nodded their heads, but I could see it felt deeply unsatisfying to many of them. After all, this meant
there wasn’t a surefire way for them to get better at product intuition and my rant on “just spend a lot of time with the data, and talk to customers, and other experts and competitors and think about what you’re thinking” felt potentially unending and fruitless.
and they just didn’t see a real-world equivalent for this.
However using AI, and maybe Large Language Models in particular, have given people a real-world equivalent that they’ve experienced that is very much as magical and unpredictable as how their brains work. :-)
You only really know what a model is capable of once you test and evaluate it
you’re still surprised by what they can do
adding more data and compute generally makes it better, though you’re not completely sure how much
I find the intuition take has been landing a lot more since I started making the AI analogy.
What do you think? Does it resonate? :- )