(San Francisco) Alphabet, the parent company of Google, published quarterly results on Tuesday that exceeded market expectations, driven in particular by the good performance of its cloud business which increasingly sells generative artificial intelligence (AI) services.
In the second quarter, the American group achieved a net profit up 28% to 23.6 billion dollars, on a turnover of 84.74 billion (+14%).
Only YouTube’s revenue came in slightly below analysts’ forecasts, at $8.66 billion.
But the market is most interested in the tech giant’s investments in generative AI, and its ability to generate revenue from it as quickly as possible.
“Our strong performance this quarter demonstrates the strength of [l’activité] “We are driving innovation across our online search and cloud journey. We are innovating across all layers of AI infrastructure,” Alphabet CEO Sundar Pichai said in a statement.
Since the release of ChatGPT at the end of 2022, companies in the sector have been engaged in a frantic race to deploy these programs for producing texts, images and other content, on simple request in everyday language (prompt).
In April, Sundar Pichai assured that he had “clear paths towards the monetization of AI through advertising, the cloud and subscriptions.”
Its cloud activity (remote computing) appears to be the primary beneficiary of investments.
Because many organizations need the generative AI models developed by leading companies to gain productivity or develop new applications.
Google Cloud held about 10% of global market share at the end of 2023, behind industry pioneer Amazon (31%) and Microsoft (26%), according to Canalys.
The spring launch of AI Overviews, Google’s search engine powered by generative AI, is not necessarily going to translate into hard cash right away, however.
Partly because it’s free, but also because Google had to correct the situation after strange, false or absurd answers to user questions.
In May, the platform suggested to an Internet user that they mix “non-toxic” glue with cheese so that it would stick better to pizza, probably inspired by comments on social networks used to train the AI model.