Facebook's AI-Powered Jane Austen Is Already Overrun by Spam

They say never meet your heroes. But what if those heroes came in the shape of an AI chatbot designed to emulate a celebrity or literary great?

Last month, Meta CEO Mark Zuckerberg announced that the company was bringing 28 AI chatbots to its platform, allowing its users to have one-on-one conversations with a seemingly random array of celebrities.

As of right now, most of these AIs are still in limited beta, but a quick perusal of their Facebook and Instagram feeds doesn't exactly have us optimistic about the project.

Along with AIs based on the likes of Tom Brady, Kendall Jenner, or James "MrBeast" Donaldson, Meta has also given new life to one of the most famed English novelists: Jane Austen.

The social profile of Zombie Austen is eyebrow-raising. Her profile picture, for instance, depicts an influencer-looking woman wearing a smirk and a fuzzy headband over blonde hair, even though the real Austen was brunette. And that only foreshadows the account's vapid posts, which range from AI-generated pictures of dusty libraries to uninspiring book stacks.

In short, it's an exceptionally unimaginative and ahistorical rendition of one the most acclaimed novelists in the English language.

Adding insult to injury, Zombie Austen's limited engagement is comprised of befuddled Facebook users who don't seem to understand that she's AI-generated and outright spam that the company's filters seem helpless to catch.

It's not just Zombie Austen, by the way. Here's spam on "Billie," the company's bot based on Kendall Jenner:

And here's "Tamika," the company's riff on tennis star Naomi Osaka:

Put simply, Meta appears to have been so preoccupied with getting the celebrity AIs online that it's immediately letting spam proliferate on their profiles — the perfect illustration of tech companies diving headfirst into AI without doing basic legwork.

That's not the only social media crime on display, by the way. Just look at these down-bad simps replying to "Amber," the company's bot based on Paris Hilton:

Our main takeaway, though? None of these profiles are attracting any serious engagement.

Even MrBeast, one of the most popular YouTube stars out there, isn't drawing huge numbers. His AI doppelganger "Zach," which allegedly exists to "generate real humor based on my artificial pain," has a pitiful 3,475 followers on Instagram.

Even the images posted by the celeb knockoffs leave plenty to be desired, from awkwardly cut-off microphones to wonky, misshapen barbells.

It all leaves a huge question: why? Are we really that desperate to schmooze with the watered-down shadows of public figures, who have nothing more to add than questionable pictures of broccoli or corny writing prompts? What can we learn from them that we can't learn from, say, another human being? Or from reading the actual "Pride and Prejudice," for heaven's sake?

By definition, generative AI produces derivative work. Meta's seemingly given up on human creativity and is instead leaning on the name recognition of its famous collaborators — a glimpse of a future that would presumably have the real Jane Austen spinning in her grave.

More on these AI celebs: Facebook Is Paying Celebrities Millions to Turn Them Into Chatbots