Remember being a teacher’s pet? Or maybe you just miss getting notes from that one professor who actually gave a damn about your prose. Well, Grammarly thinks it has the answer: simulated criticism from Stephen King, Neil deGrasse Tyson, or even Carl Sagan. Yes, the dead one.
Grammarly started as a spell-checker with ambitions. Over the years, it’s bloated into a full-blown AI writing platform. In October, CEO Shishir Mehrotra rebranded the company as Superhuman (because apparently “Grammarly” wasn’t ambitious enough), though the writing tool still goes by the old name. “When technology works everywhere, it starts to feel ordinary,” Mehrotra wrote in the press release. That’s one way to put it.
The platform now offers an AI chatbot, a paraphraser, a “humanizer” that tweaks your tone, an AI grader that scores your writing like a college course, and even tools to flag and remove phrases that sound too much like AI. Because nothing says “authentic” like using AI to hide the fact you’re using AI.
But the feature that’s raising eyebrows—and hackles—is the “expert review” option. Instead of a generic critique from a nameless LLM, you get to pick from a list of real academics and authors. Stephen King. Neil deGrasse Tyson. William Zinsser. Carl Sagan. The fine print, buried in a disclaimer: “References to experts in this product are for informational purposes only and do not indicate any affiliation with Grammarly or endorsement by those individuals or entities.”
Translation: These people have nothing to do with this. Including the dead ones, who obviously can’t consent.
Jen Dakin, senior communications manager at Superhuman, explained to WIRED that the Expert Review agent “examines the writing a user is working on… and leverages our underlying LLM to surface expert content that can help the document’s author shape their work.” She stressed that the tool “doesn’t claim endorsement or direct participation from those experts.” It just takes “inspiration” from their works.
Vanessa Heggie, an associate professor at the University of Birmingham, wasn’t buying it. She posted on LinkedIn about the feature, calling it “obscene” and accusing Superhuman of “creating little LLMs” based on “scraped work” of the living and dead, trading on “their names and reputations.” The screenshot she shared showed an AI agent modeled on David Abulafia, an English historian who died in January. Just two months dead, and already an AI ghost.
An independent review by WIRED confirmed the tool’s recommendations. It offered feedback from bots modeled on Steven Pinker, Gary Marcus, and others. For my own test, it cited “inspiration” from William Strunk Jr. and Pierre Bourdieu, “ideas” from Margaret Mitchell, and “concepts” from Virginia Tufte—all dead. Tufte died in March 2020. The advice from her AI agent: “Replace repetition with vivid, varied sentence patterns.” Riveting.
C.E. Aubin, a historian and postdoctoral fellow at Yale, told WIRED this “seems to validate the profound mistrust so many scholars in the humanities have for AI and its seemingly constant use in fundamentally unethical ways.” She added, “These are not expert reviews, because there are no ‘experts’ involved in producing them. And it’s pretty insulting to see scholarship used this way when the academic humanities are currently under attack from every possible angle.”
Beyond the ethics, there’s the practical question: does any of this actually work? Grammarly’s plagiarism detector, for instance, failed to catch a direct quotation in WIRED’s tests. So you’re getting advice from a ghost of a dead writer, but the tool can’t even spot a stolen sentence.
This isn’t just a privacy or consent issue—it’s a fundamental disrespect for the people whose work built the very models these features run on. And it’s happening in an industry that’s already drowning in lawsuits over exactly this kind of content harvesting. The fact that Grammarly—sorry, Superhuman—thought this was a good idea says a lot about where we are with AI. We’re not just automating writing; we’re automating the illusion of expertise, complete with the names and faces of people who never agreed to be part of the machine.
And that’s before we get to the dead. There’s something uniquely grim about reanimating the deceased to act as unpaid, unwitting writing coaches. It’s not just unethical—it’s tasteless. But I guess that’s the price of progress.
Comments (0)
Login Log in to comment.
Be the first to comment!