Satya Nadella insists people are using Microsoft's Copilot AI a lot | TechCrunch

"Microsoft has spent almost as much on capital expenditures in the first half of its current fiscal year as it did in all of the previous year. And the numbers truly are enormous: Microsoft spent $88.2 billion on capital expenditures last year, and has spent $72.4 billion so far this year. Much of that spend is to serve AI to enterprises and major AI labs, especially OpenAI as well as Anthropic."

"The big question on investors' minds is: will the spending turn into more use, and ultimately profits? Investors are scared that Microsoft's main enterprise cloud products, Azure, and its Microsoft 365 apps, didn't grow as fast as they wanted. "The fact that BOTH Azure and the M365 segments fell a bit short is the key negative we're hearing," Wall Street analyst for UBS, Karl Keirstead, wrote in his research note on Thursday."

"Nadella spent much of his time during the earnings call engaged in what is best described as AI use PR. Despite his pitch, some of the numbers he gave were pretty squishy. For instance, Nadella said daily users of its consumer Copilot AI products had grown "nearly 3x year over year." This refers to AI chats, the news feed, search, browsing, shopping and "integrations into the operating system.""

Microsoft reported strong quarterly results with $81.3 billion in revenue (up 17%), $38.3 billion in net income (up 21%), and over $50 billion in cloud revenue. The company has dramatically increased capital expenditures, spending $72.4 billion so far this fiscal year after $88.2 billion last year, largely to support AI services for enterprises and major AI labs such as OpenAI and Anthropic. Investors worry that Azure and Microsoft 365 did not grow as fast as expected and question when elevated spending will translate into greater usage and profits. UBS analyst Karl Keirstead flagged Azure and M365 shortfalls but still recommended buying. Satya Nadella emphasized AI adoption and cited large Copilot user growth, though some metrics were described as squishy.

Read at TechCrunch