Physical Address

304 North Cardinal St.
Dorchester Center, MA 02124

A Guide to Using AI to ‘Burn’ Users. It went Anti-Woke instead


Fable, famous social media app which describes itself as a site for “bookworms and bingewatchers,” created an AI-driven end-of-the-year summary and recap of what book users read in 2024. It was fun and interesting, but some of the recaps started again. amazing fighting tone. Briefing writer Danny Groves, for example, questioned whether he “likes to have a straight, white man’s mind” after calling him “committed to diversity.”

Author Tiana Trammell’s summary, meanwhile, concluded with the following advice: “Don’t forget the appearance of a regular white writer, okay?”

Screenshot of the reading screen on the Fable app.

Reader summary as shown on the 2024 statistics page from the Fable app.

Courtesy of Tiana Trammell

Trammell was shocked, and soon found out he wasn’t alone after sharing his experience with the Fable summary on Threads. “I received several messages,” he says, from people whose briefs contained inappropriate comments about “disability and sexual attitudes.”

From the beginning of Spotify Closedannual recaps have been taking place all over the internet, giving users a list of how many books and articles they read, music they listened to, and exercises they completed. Some companies are now using AI to create or enhance the way these metrics are presented. Spotify, for example, now offers A podcast created by AI where robots analyze your listening history and predict your life based on your interests. Legend jumped on the bandwagon by using OpenAI’s API to generate a summary of the past 12 months of reading for users, but it didn’t expect the AI ​​version to spit out a comment that took on the mien of an anti-woke pundit. .

Fable later apologized on several social media channels, including Threads and Instagram, where posted a video of an officer giving a mea culpa. “We are deeply saddened by the inconvenience caused to our Readers’ Notes this week,” the company wrote in a statement. “We will do well.”

Kimberly Marsh Allee, head of Fable’s team, told WIRED that the company is planning a number of changes to fix the AI ​​shortfall, including a way to weed out unwanted characters and a clear disclosure of AI-generated characters. “Right now, we’ve removed the part of the model that burns readers, and instead the model simply summarizes what the user likes in books,” he says.

For some users, changing the AI ​​doesn’t sound like a sufficient solution. Fantasy and romance writer AR Kaufer was horrified when he saw some sneak peeks of some of the scenes on social media. “They need to say they’re doing away with AI completely. And they need to provide a statement, not only about AI, but also an apology to those affected,” says Kaufer. “This ‘apology’ on Threads seems disingenuous, calling the app ‘playful’ as if it somehow condones racist, sexist, or homophobic comments.” In response to the incident, Kaufer decided to delete his Fable account.

So does Trammell. “The right approach would be to disable the feature and conduct thorough internal testing, including newly implemented security measures to ensure, as much as possible, that no more users of the platform are harmed,” he says.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *