The Most Surprising Conversations About Art I’ve Had This Year
I recently subscribed to ChatGPT to automate some of what I call “necessary but intensely boring busywork.”
If you had told me a year ago that some of the most interesting “conversations” I would have in 2026 would be with AI, I would have rolled on the floor laughing. I would have thought you were “off your rocker.” But that is exactly what happened.
As a working artist without gallery representation, there are never enough hours in the day. Making the work is only one part of the job. Granted, it is the part that must happen before anything else. But there I am, alone, working on my art, yet I still have a website to update, applications to write, records to maintain, opportunities to research, social media to manage, and an endless list of administrative tasks competing with my time in the studio.
I used it to draft SEO descriptions for my website; to locate dozens of links to features I had written about other artists that were still missing from my site; to make a master template containing all the information about each series, including the statement, bio, image lists, and all the necessary data about the work. These were tasks I had put off, even though I knew they needed to be done.
Somewhere in that process, something I never expected occurred: the conversations became useful for thinking about my work.
As artists, we live with our work for years. We know where projects come from. We know the experiences behind them. We know why particular images exist, why certain themes recur, and why seemingly unrelated projects are connected. But how do you know whether the thing you think you’re communicating is the thing other people are actually receiving?
As I worked through these tasks, it began drawing connections between my projects. They were often partially true, but also inaccurate. That led me to keep pushing back with some version of the same phrase: “No, that’s not it.”
I found it annoying that it kept telling me what my work was about. I made the work. I know what it is about.
Yet it kept interpreting it differently. For example, it often centered the connections on memory, identity, and transformation. All of those ideas are present in my work. None of them are wrong. Yet they are not the raison d’être for the work.
So I found myself constantly pushing back. Every time I explained why an interpretation focused on the wrong details, I was stating and restating what felt obvious. The more I disagreed, the more I was forced to articulate what was actually driving the work. Through these conversations, it became clear that the connections I considered self-evident were not necessarily visible to someone standing outside the work. The conversations became about productive disagreement and testing assumptions.
The value wasn’t in the areas that AI got right. It was that it often got them wrong in interesting ways, which made me realize how others might misinterpret my work, too.
That was the “aha!” moment.
Those interactions influenced everything else—from artist statements to website text to how I think about discussing my work publicly. Those moments revealed gaps between what I intended to communicate and what someone else might reasonably infer from the information available to them. As artists, we often talk about feedback in terms of praise or criticism. What these conversations with AI reminded me is that there is another kind of feedback: misunderstanding.
Misunderstanding can be frustrating, but it can also be illuminating. It reveals where assumptions diverge. It exposes what remains unspoken. It forces clarity.
In hindsight, the most useful thing these conversations gave me was a way to think about the distance between what an artist knows and what a viewer sees.
I started using ChatGPT to save time. I never expected it to become a tool for thinking about communication, interpretation, and the ways artists are understood—or misunderstood—by the people who encounter their work.
That was the real surprise.