In the wake of erasing around a trillion dollars of value from stocks earlier this week, the upstart Chinese lab DeepSeek was the elephant in the room during earnings calls from Meta and Microsoft on Wednesday.
Meta execs delivered strong results, and Microsoft beat earnings and revenue expectations but fell short on cloud revenue and a disappointing outlook. But that performance wasn’t the only question on the minds of many analysts attending the call.
For the last couple years, these companies have been spending tens of billions on chips, data centers, and other infrastructure to build out the future of generative AI. CEOs Mark Zuckerberg of Meta and Satya Nadella of Microsoft had committed to upping that number even more in the coming year. But will that still hold true in a world where a Chinese lab can seemingly build a similarly performing model with just a small fraction of the resources, then offer it for free or very inexpensively?
Hold steady: Nadella and Zuckerberg both reaffirmed that they intend to stay the course, for now, while attempting to learn lessons from DeepSeek’s efficiency measures. Meta plans capital expenditures of $60 billion to $65 billion in the coming year. Microsoft had previously projected $80 billion, and didn’t mention any changes to that figure.
“I think we’re still digesting. And there are a number of things that they have, advances that we will hope to implement in our systems,” Zuckerberg said in the call. “It’s probably too early to really have a strong opinion on what this means for the trajectory around infrastructure and [capital expenditure] and things like that.”
Keep up with the innovative tech transforming business
Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.
Nadella talked about how innovations like DeepSeek’s will eventually commoditize cheaper AI, which will in turn create more demand for AI usage and more infrastructure needs. (He previously posted a link on X to the Wikipedia article for the economic principle, “Jevons paradox” in reference to this idea.)
“DeepSeek has some real innovations,” Nadella said. “Obviously, now that all gets commoditized, and it’s going to get broadly used.”
Microsoft, which reaffirmed its commitment to backing OpenAI in the wake of the Stargate announcement, also said it would host DeepSeek’s R1 model for cloud customers.
Any word on ROI? For Meta, in particular, analysts and investors have long wondered how all its AI spending will translate into revenue around its main business: advertising.
CFO Susan Li said Meta is still focused on “consumer experience” before beginning to move further into moneymaking add-ons.
“Our initial focus for Meta AI is really about building a great consumer experience, and that’s, frankly, where all of our energies are kind of directed to right now,” Li said. “There will, I think, be pretty clear monetization opportunities here over time, including paid recommendations and including a premium offering.”
Cloudy forecast: Microsoft’s stock was down around 6% on Thursday after the company’s cloud revenue last quarter and outlook for the current quarter both came in slightly below Wall Street expectations. Meta’s stock, meanwhile, was up around 2%.