Meta-boss Mark Zuckerberg has unveiled 28 artificial intelligence “characters” that users can interact with on Facebook and Instagram. Photo: CARLOS BARRIA/REUTERS
Mark Zuckerberg has spent years fighting the perception that fewer people are using his apps. Now he hopes robots will bring them back.
At the annual Meta developers conference on Wednesday, Zuckerberg announced that people will be able to communicate with artificial intelligence chatbots through Facebook, Instagram and WhatsApp.
< p >The bots — 28 of them — feature digital avatars of celebrities such as Snoop Dogg, Paris Hilton and American football star Tom Brady, playing artificial intelligence «characters» who are experts in areas such as cooking, career advice and creative writing.
But Zuckerberg believes his bots need to be more than just glorified search engines. He wants them to be something closer to friends.
“It’s not just about answering questions,” he said. “It's about having fun and helping you do things to connect with the people around you. We thought it should be fun and familiar.»
Chatbots will be released gradually at first, but Meta is already promising more. It also plans to release tools that will allow developers to create their own.
Chatbots will not only be accessible through Meta messaging apps, but they will also have their own profiles on Facebook and Instagram.
With more than 3 billion people using Facebook's flagship Meta app, Zuckerberg doesn't have enough people to recruit, even though the possibilities of AI are endless.
Zuckerberg could use a win. At the same conference two years ago, the Facebook founder bet on a metaverse of virtual worlds, even changing the name to Meta.
The evidence so far suggests it failed — a decision made in the fog of a pandemic that underestimated the enduring appeal of face-to-face interaction.
However, Zuckerberg said Wednesday that Meta has not given up on the concept, unveiling a new virtual reality headset that could merge the physical and virtual worlds.
Mark Zuckerberg remains committed to his company's previous bet on virtual reality headsets. Photo: CARLOS BARRIA/REUTERS
«Not too far away now you'll walk into a room and there will be as many holograms — digital things that you can interact with — as there are physical objects,» he said.
Threads, the company's answer to Elon Musk's Twitter, is also stuttering after a strong start. Adam Mosseri, head of Instagram, which is owned by Meta, admitted last week that people aren't using it «as much as we'd like.» But AI is certainly opening up new possibilities. OpenAI's ChatGPT has been described as the fastest-growing application in the history of the Internet, and companies like Microsoft and Amazon are spending billions to fund artificial intelligence labs.
Despite its huge investment in Zuckerberg's Metaverse headset, Meta has also focused on artificial intelligence, employing some of the world's top researchers and making breakthroughs in areas such as language translation.
But now Zuckerberg has refocused the company on the technology, promising to implement it in “every one of our products.”
In addition to chatbots, Meta introduced artificial intelligence tools for image processing on Wednesday.
< p>This will allow Instagram users to edit photos using artificial intelligence, following instructions such as «Make it look like I'm on the beach» or create entirely new photos with prompts such as «Imagine a dinosaur as a pet.»
>
The CEO will be well aware of the reputational risks associated with the transition to artificial intelligence.
Many of the problems associated with AI—misinformation, election fraud, hate speech—are problems that Facebook has been accused of fueling for years. A chatbot spreading conspiracy theories would be his worst nightmare.
In fact, the company is better prepared to address these challenges than its competitors such as ChatGPT and Google Bard. Years of pressure to combat harmful content meant Meta designed its bots to be safe from the start.
“One of the benefits, for better or worse, of having all this experience in social media moderation is that you have AIs that we've been training for years to understand exactly where to draw the line,” Chris Cox , head of Meta, said the chief product officer who oversees the company's artificial intelligence efforts.
Adam Mosseri, head of another recent Meta startup, Threads, said people are using the platform «not as often as we would like to» Photo: Heathcliff O'Malley
The bigger risk is that the proliferation of bots and AI-generated images is exacerbating a trend that critics have complained about for years.
This means that Facebook and other apps are becoming less friendly and authentic: content that was once posted by friends and family is now being replaced by professional accounts designed to collect clicks.
Instagram has already applied an airbrushed filter to reality. What happens to it when your photo at the pub turns into an AI-distorted photo of you drinking a cocktail in the Maldives? What would happen to WhatsApp if bots became better conversationalists than your friends—ones that always respond to messages immediately?
Cox acknowledged these concerns. “I started working in a company when there was only text, then there were photos, there were mobile devices and video. At every step of the way, the question arises: “Is professionalization going to take over?” [because] people want authenticity.
“There is a chance that, like any other technology, it will unleash creativity. The bearish case is that we haven't yet seen this as a broad everyday use case for people.»
Zuckerberg's major recent initiatives may have failed, but Meta's lucrative business remains insulated from them. Facebook and Instagram aren't suffering because the metaverse hasn't taken off. But this latest attempt may be different.
If the head of Meta intends to flood his social networks with artificial intelligence, then robots may be the only ones left.
Свежие комментарии