‘The feature was not a good feature’ — Grammarly CEO admits Experts Review didn’t work, but you…

Grammarly CEO Shishir Mehrotra publicly acknowledged that the recently discontinued 'Experts Review' feature, which used AI to mimic writing styles of experts, was a failure, stating it was “not good for experts, it wasn’t good for users.” This admission comes amidst a class action lawsuit concerning the use of experts’ names and personas without consent. Mehrotra defended Superhuman’s (Grammarly’s parent company) approach by comparing it to platforms like YouTube, suggesting experts should have the option to build monetizable AI personas and receive compensation for their contributions.

Despite the flawed execution of 'Expert Review,' Grammarly intends to continue leveraging AI to assist experts in creating these personas. The company’s shift towards AI was driven by the increasing popularity of tools like ChatGPT and Gemini, which Grammarly saw as a threat to its utility. The author, Lance Ulanoff, a long-time Grammarly user, notes a decline in the helpfulness of Grammarly’s suggestions as AI integration increased.

Interestingly, the author’s experiment with Gemini revealed that AI platforms can already convincingly mimic writing styles of known individuals without permission or compensation, raising questions about the broader implications of AI persona usage and the potential need for legal precedents. While the outcome of the current lawsuit remains uncertain, Grammarly’s experience highlights the ethical and practical challenges of integrating AI with expert knowledge and the importance of obtaining consent and establishing fair compensation models.

(Source：TechRadar)