Paul Graham denounces the misleading use of AI in emails

Paul Graham Critiques the Use of AI in Emails

Paul Graham, an influential figure in the startup world, recently expressed his skepticism regarding the use of artificial intelligence to draft emails. He notes that many messages he receives from founders now adopt a striking journalistic style, which he believes betrays the involvement of AI. Graham asserts that no founder wrote this way before, and once he identifies a text as AI-generated, it becomes difficult for him to ignore it.

Graham confides that he has never consciously finished reading an email written by AI but signed by a human. For him, it feels deceptive, a situation he deems unacceptable. He believes it reflects poorly on the author, suggesting that they cannot write well without assistance or think they cannot do so. According to Graham, there is nothing impressive about using AI to draft texts, emphasizing that even a teenager could do it.

Reflections on AI and the Market

In addition to his critiques of AI usage in communication, Paul Graham shares other reflections on the tech sector. He describes version Claude Opus 4.8 as a "modest but tangible improvement," an observation dated May 28, 2026. Furthermore, he expresses his opinion that Anthropic and OpenAI have successfully found a "product-market fit," a remark made on May 27, 2026.

Finally, Graham mentions notes on Pope Leo XIV's encyclical regarding AI, dated May 25, 2026. These observations were collected by Simon Willison and published on May 26, 2026.