The rise of artificial intelligence in writing workflows is reshaping even the most academic professions, and not everyone is thrilled about it. A recent viral post by an American economist has sparked a wider conversation on how AI is quietly changing the nature of intellectual work.
Take a look at the X post
Marshall Steinbaum, who holds a PhD in economics from the University of Chicago, took to X to share a rather unexpected part of his daily routine. Instead of focusing purely on research or analysis, he revealed that he now spends a significant amount of time editing AI-generated text to make it appear more “human.”
Highlighting his frustration, Steinbaum wrote, “I have a PhD in economics from the University of Chicago and my main work task these days is removing em-dashes from Claude output so it’s not overly obvious.”
His remark quickly gained traction online, as many professionals relate with navigating similar challenges in the AI era. Users pointed out the irony of having to “de-AI” content, especially when the stylistic elements being removed—like em-dashes—are common in well-written human prose.
One user noted that the situation reflects a deeper issue: people are altering legitimate writing styles simply to avoid being mistaken for using AI, even though such punctuation is widely used in quality non-fiction writing.
Others responded with humor and criticism. One commenter took a jab at academic economics, joking that a Chicago PhD’s work often involves dressing up opinions with complex math hidden in appendices. Another user, curious about the editing process, asked whether em-dashes were being replaced with commas, hyphens, or split into separate sentences.
The conversation wasn’t limited to economists. A professional with a PhD in electrical engineering chimed in, saying they too spend time tweaking AI-generated corporate documents and presentations, hinting that this might be becoming a common, if strange, new normal.
Amid the banter, another user reflected on how AI is accelerating the production of ideas, albeit sometimes shaped by built-in biases and templated outputs, raising broader questions about originality and authenticity in modern writing.