Silicon Valley productivity behemoth Superhuman (formerly Grammarly, probably soon to be “MetaAIPlusGo”) has doubled down on its commitment to personalization by rolling out ‘Unauthorized Expert Mode’—a revolutionary feature that lets users get advice from AI clones of any living or deceased expert, preferably without their consent.

The feature, which was quietly introduced last August and swiftly replaced with a “please send an email if you’re upset” opt-out system, was reportedly inspired by user demand. “People are constantly saying, ‘What would Julia Angwin do?’ or ‘I wish bell hooks could proofread my LinkedIn post,'” explained CEO Shishir Mehrotra as he vigorously denied that class-action lawsuits had anything to do with the feature’s demise. “Our internal motto is, ‘It’s easier to ask forgiveness than permission—especially if you automate both.'”

Superhuman insists that the experience is highly customizable. “Expert Review used the latest in Large Language Model technology to generate advice from people you admire, people you despise, and people you’ve never heard of,” explained Lead Product Manager Chad Byte. “We saw it as democratizing attribution. Now, anyone can claim to be inspired by anyone, for any reason, with plausible deniability baked in.”

Industry insiders are impressed. “It’s the natural evolution of AI,” said Futurist and Guy With a Podcast, Trent Balmer. “Soon, everyone will have an infinite swarm of unauthorized personal mentors. Who needs real experts when synthetic ones offer bolder, more litigious advice?”

Insiders say Superhuman’s next update will allow users to direct message Nikola Tesla for startup fundraising tips, provided you sign a 70/30 revenue split.