Y Combinator founder Paul Graham says AI-written founder emails feel like being lied to
Key Points
- Y-Combinator founder Paul Graham refuses to read AI-written emails, viewing the use of text generators in personal communication as deceptive and a clear signal that the sender didn’t care enough to write the message themselves.
- Graham is likely not alone with this stance: research from Ohio State University finds that recipients frequently perceive AI-generated messages as a sign of laziness and a lack of sincerity, which erodes trust between sender and recipient.
- A BetterUp Labs survey reveals that 40 percent of US employees regularly receive low-quality AI-generated content from colleagues, with roughly half of respondents rating the senders as less creative, competent, and trustworthy.
Y Combinator founder Paul Graham doesn’t read emails written by AI. More and more founders are writing to him in a “hard-hitting journalistic style,” Graham says.
“Once you realize something was written by AI, it’s hard not to ignore it,” Graham writes on X. He says he’s never finished reading an email that carried a human’s name but was clearly AI-generated. “It feels like being lied to.”
AI should be used, but in the right way, Graham says. If not, “it makes me think less of the author. It means they can’t write well unaided (or feel they can’t), and that they’re trying to trick me.” Having AI write for you isn’t impressive, Graham adds. “Any teenager can do that.”
Graham surely doesn’t have a general aversion to AI. Y Combinator was one of the early investors in OpenAI and is deeply involved in AI investments. So why does such a straightforward use case bother him? He isn’t even criticizing the quality of the content. He clearly never gets that far.
Careless AI use erodes trust
Graham is likely running into two phenomena that researchers have also studied: social devaluation and loss of trust.
A study from Ohio State University with 208 participants found that recipients rate AI-generated messages more negatively. The reason: the sender handed the work to a machine instead of making the effort themselves.
Lead author Bingjie Liu says AI use comes across as lazy and insincere. Recipients felt less secure and less satisfied in their relationship with the sender. Liu also suspects that people now run an unconscious “Turing test in their heads,” automatically scanning messages for AI patterns.
This lines up with what Graham describes. He spots the AI style right away because no founder has ever written to him that way before. Once the AI use is detected, attention shifts from the content of the message to a different question: how little does this sender actually care about the recipient?
Researchers from BetterUp Labs, working with Stanford’s Social Media Lab, surveyed 1,150 US employees. According to the results, 40 percent regularly receive shallow, substance-free AI content from coworkers.
The social consequences are clear. Fifty-three percent of respondents said they feel annoyed by such content. About half rate workslop senders as less creative, less capable, and less reliable. Forty-two percent consider them less trustworthy. A third said they want to work less with these people going forward.
The BetterUp Labs researchers distinguish between “pilots,” who use AI with purpose and autonomy, and “passengers,” who rely on it mainly to avoid work. Graham apparently places founders who send AI emails squarely in the second group.
Subscribe now
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.
