Grammarly’s New AI Lets Authors ‘Review’ Your Writing
▼ Summary
– Grammarly has expanded from a grammar tool into a comprehensive AI writing platform, now part of the rebranded parent company Superhuman.
– Its new features include an AI chatbot, style paraphraser, voice humanizer, and a tool to flag AI-generated phrasing.
– A controversial “expert review” feature simulates critiques from real-life authors and academics, both living and deceased, without their permission or endorsement.
– The AI agents are likely trained on the works of these experts, raising legal and ethical questions about copyright and the use of their names and reputations.
– Critics, including academics, have condemned the practice as obscene, accusing the company of creating AI models from scraped work and exploiting these figures’ legacies.
Imagine having a personal writing coach who can channel the wisdom of literary giants and celebrated scholars. That’s the ambitious promise behind Grammarly’s latest AI feature, which allows users to receive feedback inspired by real-world experts. This tool, part of the company’s broader rebranding to Superhuman, aims to provide a more personalized and authoritative critique than a standard grammar check. However, this innovation raises significant questions about intellectual property and the ethical use of a person’s legacy, as the system leverages the published works of individuals, both living and deceased, without their direct consent or involvement.
The platform has evolved far beyond its roots in correcting spelling and punctuation. It now offers a comprehensive suite of AI-driven tools designed for virtually any writing scenario. Users can access a chatbot for drafting assistance, a paraphraser to adjust style, and a humanizer to revise text according to a selected voice. There’s even an AI grader that estimates how a document might score as academic work and tools to identify phrases that sound overly AI-generated. The underlying goal is to make advanced writing assistance feel seamless and integrated.
The most controversial addition is the expert review option. This feature presents users with a list of notable academics and authors, such as Stephen King or Neil deGrasse Tyson, whose perspectives can ostensibly guide revisions. A critical disclaimer notes that these references are for informational purposes only and do not imply any affiliation or endorsement from the individuals named. The system also includes figures who have passed away, like editor William Zinsser and astronomer Carl Sagan.
According to the company, the AI agents are designed to surface relevant expert content based on the user’s writing. A senior communications manager explained that the tool examines the document’s substance and leverages the company’s large language model to provide suggestions inspired by the works of these experts, directing users toward influential scholarly voices. It does not claim direct participation from the experts themselves.
This approach has sparked considerable debate, particularly concerning copyright and ethical boundaries. Critics argue the company is effectively creating specialized AI models trained on scraped content from published works, commercially utilizing the names and reputations of individuals without permission. A recent example highlighted by a university professor involved an AI agent modeled on a historian who died earlier this year, a practice she labeled as obscene. The legal landscape surrounding such use of copyrighted material to train AI remains unclear and is actively being contested in numerous lawsuits.
While the technology represents a significant leap in automated writing assistance, it forces a confrontation with complex issues of ownership and legacy. The feature offers a novel way to engage with expert ideas, but it does so by operating in a legal and ethical gray area, leveraging the intellectual output of people who may have no say in how their life’s work is algorithmically repurposed.
(Source: Wired)
