(San Francisco) Meta, the parent company of Facebook, Instagram and WhatsApp, doubled its net profit in the first quarter, but the increasingly high cost of its investments in artificial intelligence (AI) worries Wall Street.
The world’s number two online advertising company saw its quarterly turnover increase by 27%, reaching $36.5 billion, of which it generated $12.4 billion in profits, both results above expectations.
But Meta also announced on Wednesday that its investments would be between $35 and $40 billion this year, more than expected, because of the needs in AI (infrastructure, research and development, etc.).
Information which stunned the market: the Californian group’s shares lost more than 16% during electronic trading after the close of the New York Stock Exchange.
“I believe we have gained in optimism and ambition on AI,” Mark Zuckerberg, the company’s boss, tried to explain during the conference for analysts.
Lagging behind Google and Microsoft in the frantic race for generative AI (production of texts, images and other content, upon simple request in everyday language), Meta unveiled last week the new version of Meta AI, its assistant which responds to user questions, like ChatGPT.
Having remained very discreet until now, Meta AI will gain visibility on the group’s platforms and skills thanks to a new, more advanced AI model, Llama 3.
“First AI company in the world”
Mark Zuckerberg assured that the first returns were very positive, and encouraged him to invest sufficiently to “stay at the cutting edge” and make Meta AI “the best and most used AI assistant in the world”.
“We’re no longer just training good AI models to build new products for social media and e-commerce,” he insisted.
“We are at a point where we are showing that we can build cutting-edge models and become the leading AI company in the world.”
However, he acknowledged that it will undoubtedly take “several years” before these efforts bear fruit.
Too long and too uncertain for Wall Street. During the conference, some analysts even suggested that Meta spend less in the metaverse and thus free up funds for AI.
The Reality Labs branch, responsible for developing devices and software for the metaverse (mixing real and virtual universes via high-tech glasses and headsets), once again recorded significant losses, of more than $3.8 billion. .
And the company predicts that they will widen further.
But Mark Zuckerberg, who considers the metaverse to be the future of the internet, reiterated his confidence in the potential of these technologies.
Especially since AI goes well with these new objects, according to him: “Connected glasses are ideal for the AI assistant, because they allow it to see what you see and hear what you hear. So the assistant has all the context to help you.”
“Enviable margins”
The American group has just launched “Meta Horizon OS”, its operating system for mixed reality devices (combining real elements with augmented and virtual reality), now open to other manufacturers.
Some analysts are more optimistic, given the growth of its core business.
Meta has more and more advertising space available for sale, at an increasing average price, enough to “generate enviable margins, even if it continues to invest in sectors which may not contribute to profits before several years,” reacted Max Willens of Emarketer.
By the end of the year, Meta could also start selling advertising on Threads, its written message platform similar to X (formerly Twitter).
This would please “advertisers who want to communicate in real time with their audience and who will finally have a viable alternative to X,” noted Mike Proulx, vice-president of Forrester.
In AI, Meta has fallen behind, agrees Debra Williamson of Sonata Insights.
But “thanks to its platforms, it has a massive user base for testing AI experiments […] and quickly assess those towards which its users gravitate,” she stressed.
The company identifies 3.24 billion people who use at least one of its services on a daily basis, or 50 million more than at the end of 2023.