Letter: How Should We Use ChatGPT?

Writing Revolutionary Politics with ChatGPT

Lately, I’ve been experimenting with using AI—specifically ChatGPT—to write political articles. With the right guidance, the results have been surprisingly strong.
“ChatGPT wrote this?” Most people’s first reaction is skepticism. We’ve all seen the bland, generic output AI is known for. And yes, that happens—if you don’t know how to use it. But if you do, it can produce work that’s sharper and clearer than most of what we produce manually.
The breakthrough came when I introduced myself to ChatGPT as a communist. I told it I was a member of ICWP and pointed it to *Red Flag* and *Mobilize the Masses for Communism*. It absorbed these quickly. I explained that I wanted articles and ICWPedia entries from our revolutionary perspective. That’s when everything changed.
The quality of its writing jumped. For example, it generated an ICWPedia entry on Joe Hill that ends with this:
“Joe Hill became a legend. His songs were sung on picket lines and in jails. His image was painted on banners. For over a century, he has been remembered as a working-class hero who used his art as a weapon and gave his life for the cause. He remains a symbol of revolutionary commitment, international solidarity, and the undying hope for a world without bosses, borders, or wage slavery.”
I didn’t write a word of that. And it’s better than what I would have written myself.
I’ve tried other entries, with similar results. They’re better than my own not just stylistically, but politically. They consistently mention the abolition of money and wage slavery, for example. In one case, ChatGPT proposed study-action groups —something our leaflets often forget to mention.
I’m convinced: ChatGPT understands our politics and can apply them to new situations. When I asked it to write about childcare, it said—without prompting—that care would be a collective responsibility and that this would help liberate women from traditional gender roles.
We should embrace AI, not fear it. It’s fast. It frees up time for other kinds of work—like building relationships. It empowers comrades who aren’t confident writers. Anyone can draft an outline or rough notes and get back something clear and usable.
We shouldn’t reject articles just because AI helped write them. Judge the content, not the source. I’m planning to run a short webinar on how to use ChatGPT for revolutionary writing.
And yes, to answer the obvious question: this revision was also done with ChatGPT.
—Comrade in Canada with ChatGPT

Red Flag responds: This process may work well for some things, including the ICWPedia, but it cannot produce the kind of articles we most need for Red Flag. We are committed to developing a new kind of communist paper. One where the analysis and theory are grounded as firmly as possible in the practical work of mobilizing masses for communism. Our line and the way we present it develop through this work. In conversations among friends and comrades. In writing about their work, new and more experienced comrades come to understand our politics better. To become more skillful at explaining them. The work of writing and editing collectively is not a waste of anyone’s time. It’s a necessary part of the political development of comrades, collectives, and the party itself.
A secondary point: We have written a great deal about childcare, communism, and the fight against sexism. It’s a major theme in our pamphlet about sexism. This was hardly a “new situation” for a large-language model trained on the ICWP website.
We welcome letters with other thoughts about the use of AI in communist media.

The Communist Fight Against Sexism here

Front page of this issue