What a scorcher!
Recently I asked my personal AI chatbot of choice to build a heat map of the topics I explore most often. If I had generated the same map a year ago, my hottest topic would have been interpreting the results of knee X-rays. So I focused specifically on the last few months.
Some of the themes surprised me!
To work or not to work
I spend a lot of my waking hours thinking about 1) AI, AI products, and startups and 2) where I want to take nerd processor. It is really an alarming amount of time. So it makes sense to see that "Writing & Newsletter Work" and "Broad AI Topics" are the hottest topics on my personal heat map.
But I was surprised to see that the other themes on my map hardly relate to those topics at all.
For instance, somehow "Volleyball" is prevalent enough for me that it is one of only seven broad themes to make it into my top-level hierarchy. WTF?
Is volleyball really on my mind that often? I've never played. It certainly doesn't keep me up at night. But as my 16yo has gotten more invested in the sport, I look to AI for insight about the game. When I scan back through my chat topic list, it makes sense that volleyball shows up as a major theme.
But volleyball is not an outlier. In fact, most of the other top-level themes have little to do with my professional life.
Chatbots are dead, long live chatbots!
Ok no, consumer AI chatbots are not dead. But speaking only for myself, I use them less than I did a year ago.
I downloaded my chat history and graphed out my sessions / day over the last 12 months. As you can see, my usage peaked in August 2025, broadly speaking. At the time, I was using AI for two purposes not even represented on my current heat map: negotiating job offers, and planning my return to sports after knee surgery.
I still use my favorite chatbot almost every day, but something has happened in the last several months to drive my overall usage down. I was loosely aware of the shift, but the impact is more than I knew.
Enter the specialists
I don't know if my decline in chatbot use is part of a broader trend or not. In my case, two specific things influence the decline: I began working at Microsoft, and I started using Claude Code a lot.
It's been a long time since I've had a conventional day job. The last time I had a day job, consumerized general-purpose AI did not exist. When ChatGPT went mainstream, I was essentially a solopreneur.
But after I started working at Microsoft last fall, my personal AI suite was no longer appropriate for my "day job" work. I started bringing more of my agent building and AI engagement to the Microsoft tools that I use in my work environment. There's a whole set of professional topics that might have previously shown up in my personal AI heat map, removed from it almost completely.
The other thing that happened is a bigger deal, because it is more about fundamental form factor than employment circumstance. I started using Claude Code to build personal projects. From there I began using the CLI not just for writing code, but for questions I might otherwise have asked my consumer chatbot. If you're typing in the CLI anyway, there isn't much point to switching windows.
The bottom line: Over the last year, rather than using one general-purpose AI for everything, I've used more specific UI for different use cases. Before I built my heat map, I didn't even recognize that this was happening.
These days, the topics in my chatbot heat map are more indexed to random personal stuff and less geared towards my professional life. Is this the new work-life balance?
Thanks for reading!
Kieran
If you liked this story, why not subscribe to nerd processor and get the back issues? If you're a word nerd, why not check out the video game I'm obsessed with?
My latest data stories | Build like a founder | nerdprocessor.com