The Facebook Group Meta brings chatbots and other functions with artificial intelligence into its apps. The software will initially be available in the USA. Meta joins other tech heavyweights who are upgrading their AI offerings following the success of the chatbot ChatGPT.
On Wednesday, Meta boss Mark Zuckerberg presented, among other things, software that can quickly generate images from text specifications, which can then be used, for example, in posts on online networks. On the other hand, Meta wants to let users interact with AI bots that specialize in different topics and areas of knowledge. Celebrities lend their faces – such as football star Tom Brady for a bot on sports topics.
The idea behind it is to give users an opportunity to pursue their interests in an interactive way, said Meta product boss Chris Cox. In the “AI Studio” companies should be able to develop their own artificial intelligence. Celebrities could also bring AI versions of themselves to the platform over time, Cox said.
Artificial intelligence will change how people use most of Meta’s products, emphasized Zuckerberg. The group also includes Instagram and the chat services WhatsApp and Messenger. Because AI software can sometimes produce unexpected and problematic content, Meta will take more time to introduce such features than usual.
At the same time, Cox admitted that for most people, AI applications have not yet become an integral part of their everyday lives. “They’re definitely interesting, but we haven’t seen anything that would be useful on a day-to-day basis like WhatsApp, email or the camera.” But he could imagine that artificial intelligence could be very helpful, for example, when learning new languages.
Zuckerberg also presented new VR glasses that are intended to link the real environment and digital content more closely. The device called Quest 3 uses cameras to record the room and displays it in three dimensions. Digital objects can then be embedded in the real environment for users. This could be, for example, virtual photos on the wall – or a portal through which you can slip into another virtual world.
Zuckerberg said he assumes that over time you will interact in a room with as many digital objects as real ones. During consultations in companies, for example, there are increasingly people, digital avatars of absent employees and various chatbots sitting at the same table. Everyday life increasingly consists of a combination of physical and digital experiences and a central question is how to bring them together.
Zuckerberg has been relying on virtual worlds for a long time. He had the Facebook group renamed Meta to emphasize the focus on the virtual environment, the “Metaverse”. At the same time, the development is consuming billions of dollars quarter after quarter – while the technology has still not conquered a mass market.
In the first half of 2023 alone, the Reality Labs division posted an operating loss of $7.7 billion. However, technical boss Andrew Bosworth was convinced that investments would pay off in the end. “You can see from the numbers in the quarterly reports how certain we are about this,” he told the German Press Agency.
The Quest 3 also brings the VR pioneer Meta into direct competition with Apple, which also wants to merge real and artificial worlds with its soon-to-be-released Vision Pro glasses. However, Meta emphasizes that they want to make the technology more affordable: Quest 3 costs $499 in the USA, while the Apple glasses will come onto the market next year for $3,500.
Meta also presented a new version of the smart glasses developed together with eyewear specialist Ray-Ban. It can take pictures and videos – and the AI chatbot Meta AI can answer users’ questions via tiny speakers. For example, you can ask the glasses what kind of building you are currently looking at.
If you ask the glasses to send a photo to a friend, they will take a picture and send it via your usual messaging app. If desired, the software can also automatically write text to an Instagram video you have just recorded.
Meta can keep risks that AI brings under control, said Bosworth. It is not difficult for users with malicious intentions to control an invasive chatbot. “But we still control accounts’ access to the platform.”(dpa)