The characters of Meta AI are live, and they are just as scary and confusing as you could expect.
In Meta Connect 2023, CEO Mark Zuckerberg presented A collection of various AI characters with unique interests and personalities. Thanks to Facebook Messenger, Instagram and WhatsApp, users can chat with specific characters on subjects such as cooking, training sessions, travel or general life advice. THE tool is now in live beta mode, giving users a chance to try it for themselves.
It is fundamentally the same concept as chatbots as Replika, which allows users to design and interact with their own IA companions. Except with the Meta version, many chatbots are based on real famous people. Celebrities look like themselves, but they are not themselves, they play characters.
Confused? Here are some examples: Paris Hilton plays Amber, a detective who helps you to solve crimes, Kendall Jenner plays Billie, a type of older sister who gives life advice, Duane Wayde is Victor, an Ironman triathlete who helps you motivate yourself, and Tom Brady like Bru, who likes to discuss sports. Why this confusing concept of celebrities playing someone else? Is it to prevent Deepfakes? Protect from responsibility? If so, why use celebrities in the first place? Despite this, the spokesperson for celebrity Gimmick does not inspire much confidence.
In case you are wondering how Meta forced celebrities to sign their resemblance and launch their AI replicas at the mercy of the meta-users, the answer is millions of dollars. According to InformationCelebrities and creators were paid up to $ 5 million for the use of their resemblance, which involved up to six hours of studio time to record reactions and poses.
Mashable lighting speed
The most clumsy conversation of all time
Celebrity videos do not end up playing an important role in experience. Despite being presented as a humanizing interaction with AI, the actual experience is simply to send SMS with a chatbot. Studio hours were intended for a small window at the top of the screen that shows the characters reacting, smiling and posing. Disappointing, this aspect does not correspond to your conversation. When I used the suggested prompt “I need to let off steam”, it coincided with the video of Kendall Jenner, I mean Billie, stifle the laughter and turn my eyes without interest. Not so favorable to a big sister.
Billie does not seem super interested to let me evaluate.
Credit: Meta
Aside from the emotional wound, it could cause people who really need to let off steam, I expected more technologically. Why can one of the largest social media societies in the world not react the characters in AI? My intuition is that this can, but the functionality is intentionally nerve-ed to prevent the use of sequences for Deepfakes. But this kind of cancellation of the whole sale of being able to interact with celebrities, because it really seems to be an AI chatbot with a pre -recorded video launched on the top.
It’s AI, acting like real people – but badly
Even if Meta obtained the right contextual reaction, there is something deeply disturbing in these characters. With Chatgpt, I know I am talking to a bot, and it doesn’t try to be something else. Same replickas, which humans have formed real, and sometimes inappropriate obligations With, look at animated, creating a critical border between the real and artificial world. But to discuss with the characters AI of Meta is as if the tool tries to convince me that the characters are human and real.
Like Jules Terpak noted In a video on X (formerly known as Twitter), bots are “used as companions to lower you”. Even when Terpak says: “Bye” to the character of MrBeast Zach, he answers: “No question of guy, we are just starting …” In a somewhat desperate attempt to continue to speak to him.
Tweet may have been deleted
As for most social media platforms, it is no secret that Zuckerberg’s strategy for Meta applications is Keep users on the platform During as long as possible – generally because it brings more advertising revenue.
Personas is a more accessible chatppt, just with the faces of Celebrity
Aside from goosebumps, I could see this working for some people – especially with characters who have specific expertise such as cooking or exercise. As Replika proves, people are arranged and capable of developing relationships with AI companions, and the Meta tool has the advantage of being easily accessible to anyone with Facebook, Instagram or Whatsapp. But I end up with persistent questions about confidentiality and data. Unlike Chatgpt, there is no obvious way to clean the cat’s history. To do this, you need to type / reset to the cat to delete the copy of the AI of your messages. Or, if you want to delete copies from your messages with all AI, type / reset everyone. In addition, in Instagram settings, the The option to use end -to -end encryption is grayBy saying “some people cannot use it yet.”
Max gave me recipe ideas, but was not more useful than other chatbots.
Credit: Meta
The most useful experience I had was with Max, played by the famous Chef Roy Choi. Like the other characters I discussed with, Max’s reactions had nothing to do with what I was asking, so I especially ignored the window. With this out of the equation, it was a standard AI chatbot that gave me ideas and fall soup recipes. At this point, I simply used the tool as a chatgpt, which made me ask me why not just use chatgpt. If Meta tries to keep users on their applications, they will need more than celebrities.
Subjects
Artificial intelligence
Meta