John Birmingham Using AI To Write Pisses Me Off

2nd June, 2025

John Birmingham's opinion piece in Six Colors begins as a straightforward story about a guy who finds a writing tool that works really well for him, but by the end of the piece, the treachery on display crossed my imaginary picket line in the battle between creatives and artificial intelligence.

The piece is titled "Whisper is an AI-powered jet engine for writing" and John isn't wrong. Whisper is a huge improvement over other dictation software and for many writers, dictation is the preferred way to pump out thousands of words a day without copping a heavy dose of repetitive strain injury. Good for him, I'm glad he likes it. What boiled my piss is the last couple of paragraphs:

"there's a second step to getting a good, clean copy out of a dictation rig like MacWhisper Pro: You have to feed the transcript to an AI like Claude or ChatGPT and ask it to clean it up for you"

John is giving the output from Whisper to another service to "clean it up for you". Yes, dictation does need to be cleaned up but that is your job as a writer. I assume John edits it himself after ChatGPT does - so why bother giving it to ChatGPT in the first place? Do not trust the machine to "interpret what I said, but what I meant", do it yourself! Chuck the text into Word, read it, and edit it as you go, applying your personality, your thoughts, your humanity into whatever it is you are writing and regardless of who is paying for it.

"Part of the reason these models are so good at accurately interpreting not just what I said, but what I meant, is that they have consumed every word I ever published. They did it without my permission, and the billionaires who own and run these companies say they can't possibly afford to pay for any of it."

John admits that his work was stolen, which I think most writers have come to terms with now because that boat has sailed, but what angers me is that John assumes because ChatGPT has "consumed every word I ever published" the machine is "so good at accurately interpreting not just what I said, but what I meant".

Is John for real? It's bizarre to read something so stupid from someone normally so astute. Why would you outsource the most critical part of writing, your intent and your meaning, to a machine, because you think it understands you? The machine will never understand you. It's a goddamn machine, made by humans that hate what you do for a living and wish to commoditise your craft.

How do the 7,000+ subscribers of John's Substack feel about this process of having ChatGPT "accurately interpreting not just what I said, but what I meant" in the writing they pay $130 a year for? When I buy a book I don't want to know what a computer is thinking or feeling because it is something a computer is incapable of.

John admits in the piece that what he wrote - did he even write it? Did he get ChatGPT to write it for him and he just slapped his name on it? Who knows?? - is controversial with writers, but did it anyway and it sounds to me like he is proud of it. Proud that he found a writing hack to let him get more work done in a smaller amount of time, just like the good little productive nodes the machine wants us to be. Congratulations.

(no LLM or AI was used to dictate, type, edit or publish this post)