AI1 is ruining writing.

I don’t mean how AI slop is polluting the internet. I don’t mean how our reliance on them is degrading our skills. I mean flesh and blood writers avoiding certain patterns to duck AI-use accusations. The most common one: the em dash.

The New York Times defended the em dash better than me. My gripe is the unexpected burden AI has put on humans — proving they are human.

We all despise reCAPTCHA, but generative AI and LLMs2 are a different beast. Job candidates are holding up a hand in front of their face in zoom calls. Students run their essays through AI detection algorithms before submission.

But I have no one to convince. Even if it is a GPT3-ism, IDGAF4 . I will continue to use my beloved em dash until data centers cause the heat death of humanity.

Why are there footnotes?

  1. artificial intelligence ↩︎

  2. large language models; Chat-GPT, Gemini, Perplexity, etc. are examples ↩︎

  3. Chat-GPT ↩︎

  4. I don’t give a f*ck ↩︎