Commentary

Pretend Emails: Y Combinator Co-Founder Decries AI-Written Messages

Business owners—copywriters, too—should heed a recent comment by Paul Graham, co-founder of Y Combinator, on email and artificial intelligence.

Graham says he stops reading when he detects an email was written with AI. That's easy to figure out, he says.

“A lot of the emails I get from founders are now written in a hard-hitting journalistic style,” Graham wrote on X. “I know they’re written by AI, because no founder ever wrote this way before. And once you realize something is written by AI, it’s hard not to ignore it.”

Agreed. But the emails, articles and other AI-written things I’ve seen do not reflect the polished prose or journalistic style Graham mentions. Rather, the writing tends to be stilted, meandering and often factually inaccurate.

But Graham has a point. “I have never knowingly finished reading an email signed by a human but written by AI,” he writes. “It feels like being lied to, and who would stand for that?”

advertisement

advertisement

Graham adds that AI-written emails do not show efficiency. And that’s true of marketing emails as well as the outreach messages coming from a CEO. 

He also decries machine-written emails that mimic a human voice.

Graham should know. But here’s another knowledgeable source: OpenAI’s Sam Altman.

“I’ve felt that threshold -- most strongly when I briefly tried to let AI do my messaging,” he said, speaking in Australia, according to Startup Daily. “You know, write my emails, my text messages, do Slack for me. Yet, I found it, like, surprisingly dehumanizing to watch, even when I had it reply to messages.”

As the great direct-mail copywriter Bill Jayme said, the writing he and others did in his day “conveyed warmth and it conveyed charm. We tried to reward the reader for his reading time."

Tell that to a machine learning model. 

Next story loading loading..