Poets and Journalists
I use Gemini for coding, a lot. When I tell some colleagues, they look at me like I’m out of my mind. I believe their reaction reveals a fundamental split in how we see our craft: the difference between the Journalist and the Poet.
Ever since I started using Gemini for coding, the amount of projects I have started and finished has been amazing. Easily in the dozens. It has been a massive personal productivity boost, especially off-work. At work, I have not used it much for coding until recently, but for all the other things it can do (planning, summarizing, finding information) I use it every day, multiple times. I know some people who are several levels beyond my own usage, commanding armies of agents creating code for them. I’m not that far down, I’m just at the place I feel best about it.
Then, I talk with some colleagues and friends and they look at me as if I was some sort of lunatic. The reasons for the look are as diverse as they are. Some wonder how they could trust the code is maintenable. Some argue that they are faster coding stuff themselves than telling the LLM what to do. Others just report that they tried it once, could not get it to work and gave up.
I believe this divide isn’t about technology, but about how we relate to our craft. Two archetypes that live within every developer: the Journalist and the Poet.
There is no value judgement in using these two words. They should just be taken as useful archetypes, and neither is better than the other. Most of us will shift from them depending on the task, day of the week or how much coffee we have drunk that day.
The Poet values the craft of coding as an end in itself. They value clean, beautiful, maintainable code they understand from side to side. For the poet, the thrill is in designing the perfect algorithm or interface, or squeezing the last bit of performance from a tricky loop. The act of writing the code is the act of thinking.
The Journalist, on the other hand, sees code as a means to an end. The goal is to deliver an experience, or solve a problem. While they appreciate the same virtues the poets seek, their primary focus is on the result. The code is the medium, not the message.
What this explains
This distinction makes our reactions to LLMs easier to understand. As someone who is most of the time a journalist, my goal is to get something done. The LLM is the ultimate tool for this: it helps me get there faster. Yes, I care about maintaining this code and extending it, but for the initial draft, boilerplate, and the first iterations until the concept is nailed the AI is a powerful tool to remove friction.
From what I have seen, those who lean more towards the poet side struggle to see the value in “getting there” if they are not the ones driving. The LLM short-circuits the creative process that they enjoy. Handing off that part likely feels like handing off thinking. This makes them adopting LLMs as helpers or assistants slower or more difficult.
Recognizing this difference helps me reframe how I talk to developer friends and colleagues. Understanding whether someone is in a poet or journalist mindset helps me know how the conversation needs to be framed, how we should talk about tools, and how we should discuss quality.
Just remember: we aim for the same destination, we enjoy different parts of the journey.
