AI-powered email assistants create perfect, personalized messages with minimal human intervention. But some people are now concerned that their emails sound too perfect, including people who work in tech.
SCOTT SIMON, HOST:
Composing emails has never been easier. AI assistants can create messages without much guidance from real people. But now some real people are worried that emails are being read too perfectly. And as NPR's Chloe Veltman reports, those concerns are shared by some who work in tech.
CHLOE VELTMAN, BYLINE: Email hacks have taken over our inboxes. AI-powered tools can do things like compose messages in response to simple prompts and complete people's sentences. AI-powered email is big business too. The robomail market is expected to grow to more than $4.5 billion by 2029. However, a recent study from Columbia University shows that more than half of spam messages are written by artificial intelligence, which is one reason why some tech workers are concerned that their AI-powered emails will be seen as such because they seem too machine-like.
PETER SLEIMAN: Sometimes I make small typos on purpose if I want to seem human.
VELTMAN: This is Peter Suleiman. He is the creative director of a video agency.
SLEIMAN: Forgetting a period or maybe misspelling a word, maybe adding an old-fashioned emoji instead of real emojis like ChatGPT does. I could add two periods in parentheses.
VELTMAN: The topic of inserting intentional errors into email came up during a session Sleiman taught on digital marketing at the recent TechCrunch Disrupt conference in San Francisco. Fellow entrepreneur Luca Oak attended the session.
LUCA OAK: One of my favorite t-shirts is the one that says “I'm Secretly Grading Your Grammar.”
VELTMAN: Oak says he prefers people to stick to good grammar and spelling, please.
OAK: Because actually, your typos can make it very difficult to understand your message when I have to dig into what you meant here?
VELTMAN: But even Oak has its own method of humanizing email.
OAK: I write all I's in lowercase.
VELTMAN: But defeating AI's desire to make perfect suggestions requires commitment.
OAK: Autocorrects me all the time.
VELTMAN: Graduate student Daniel Zock was also at the conference, learning about the latest innovations. He says people in his network are even asking AI assistants to write in less advanced ways to appear more human.
DANIEL ZOK: I say, hey, ChatGPT, please avoid the em dash and trailing dash and all that stuff.
VELTMAN: But Zock says he's personally fine with AI composing slick emails on his behalf.
ZOK: What are we trying to do? We're trying to train these systems to achieve that level of humanity. So it would be nice if the AI would generate something very close to what a human would output, wouldn't it?
VELTMAN: Zoke says people shouldn't get so hung up on style. The content of the message is important.
Chloe Weltman, NPR News.
(SOUNDBITE OF GHOSTOWN DJS SONG “MY BOO”)
© 2025 NPR. All rights reserved. Visit our website terms of Use And permissions pages in www.npr.org for more information.
The accuracy and availability of NPR transcripts may vary. Transcript may be modified to correct errors or accommodate audio updates. Audio on npr.org may be edited after it is originally broadcast or published. The authoritative recording of NPR programs is the audio recording.
