DeepSeek, DeepSeek, DeepSeek: CEOs keep getting asked about the Chinese AI upstart on earnings calls

Lisa Su, the CEO of AMD, said DeepSeek is driving innovation that's "good for AI adoption" in an earnings call on February 4.

"The fact that there are new ways to bring about training and inference capabilities with less infrastructure actually is a good thing, because it allows us to continue to deploy AI compute and broader application space and more adoption," Su said.

She added that AMD is a "big believer" in open source.

"And from that standpoint, having open source models, looking at the rate and pace of adoption there, I think, is pretty amazing," Su said. "And that is how we expect things to go."

Google's CEO Sundar Pichai said in an earnings call on February 4 that DeepSeek had done "very good work."

"Look, I think there's been a lot of observations on DeepSeek. First of all, I think a tremendous team," he said to investors. "I think they've done very, very good work."

He said that for Google, it had "always been obvious" that frontier models could be made to be more efficient over time.

But he downplayed DeepSeek's threat, saying that he thinks Google's Gemini model is the "Pareto frontier of cost, performance, and latency."

He added that Google's recent 2.0 Flash Thinking models are "some of the most efficient models out there, including comparing to DeepSeek's V3 and R1."

Shyam Sankar, CTO of Palantir Technologies, fielded a question about DeepSeek during the company's earnings call on February 3. He said DeepSeek demonstrated that AI models are "commoditizing."

"But I think the real lesson, the more profound one, is that we are at war with China," Sankar said, adding, "We are in an AI arms race."

He also criticized the explanation that "the Chinese just copy and we're the only innovators," an apparent reference to reports that DeepSeek may have copied OpenAI.

He said the engineering in DeepSeek's R1 model was "exquisite" and that "the optimizations that they've done are really impressive."

"We have to wake up with the respect for our adversary and realize that we are competing," Sankar said, adding, "We have to realize that the AI race is winner take all."

"The time to mobilize has come," he said.

An analyst asked Apple CEO Tim Cook for his perspective on the "DeepSeek situation" during the company's quarterly earnings call on January 30.

"In general, I think innovation that drives efficiency is a good thing," Cook said. "That's what you see in that model."

The CEO said he thought the company's "tight integration of silicon and software" would continue to serve them well.

"From a CapEx point of view, we've always taken a very prudent, deliberative approach to our expenditure, and we continue to leverage a hybrid model, which I think continues to serve as well," Cook said, referring to Apple's AI strategy.

Meta CEO Mark Zuckerberg acknowledged DeepSeek as a "new competitor" during an earnings call on January 29. An investor asked him about the competitive dynamic in the open-source field.

"In light of some of the recent news, you know, the new competitor, DeepSeek from China, I think it's also one of the things that we're talking about, is there's going to be an open-source standard globally, and I think for our own national advantage, it's important that it's an American standard," Zuckerberg said to investors.

He added that the emergence of DeepSeek has "only strengthened our conviction that this is the right thing for us to be focused on."

Later in the call, he said that DeepSeek did "a number of novel things" to train its model fast and cheaply, which Meta was "still digesting." He added that DeepSeek has made advances that Meta hopes to implement in its systems.

Microsoft CEO Satya Nadella mentioned DeepSeek twice in his prepared remarks during an earnings call on January 29.

He said that the Copilot+ PC laptops, which Microsoft has called the "fastest, most intelligent Windows PCs ever built," would soon be able to run DeepSeek's R1 distilled models locally.

When asked about DeepSeek by an investor, he said, "I think DeepSeek has had some real innovations. And that is some of the things that even OpenAI found in o1."

IBM's CEO Arvind Krishna fielded a DeepSeek question during an earnings call on January 29.

When asked about what implications DeepSeek could have for IBM or the industry at large, Krishna said, "Look, DeepSeek, I think, was a point of validation."

"We have been very vocal for about a year that smaller models and more reasonable training times are going to be essential for enterprise deployment of large language models," he said.

The tech giant's chief added that IBM has been going "down that journey" itself "for more than a year" and that it has seen "as much as 30 times reduction in inference costs" with those approaches.

"As other people begin to follow that route, we think that this is incredibly good for our enterprise clients," Krishna said.

AT&T CEO John Stankey said the newer, lower-cost AI "is going to open up and facilitate new applications and business models."

"This is a seminal technology cycle," Stankey said on January 27 of generative AI. "It's going to be every bit as big as the founding of the Internet when it's all said and done."

Stankey added that new breakthroughs like DeepSeek that use less processing capacity, consume less power, work more effectively in particular domains, or can be run on local devices instead of in the cloud will ultimately lead to new applications and business models.

"We're all going to have to stay on our game to make sure we use it effectively so none of us are in a disadvantaged position relative to our competitors on cost-structure effectiveness," he said.

Wendell Weeks, CEO of glassmaker Corning, which produces fiber optics that are increasingly critical in high-speed networking, said the technical community has been watching DeepSeek for the last few months.

"What's super important to understand is that we need dramatic improvement in training and inference cost to make GenAI into a highly sustainable business model, and more importantly, the productivity driver that we all hope it will be," he said.

"All of us in the space are counting on many more innovations to come," he continued, adding that AI models of the future will continue to need improvements in computation and communication technologies.

Revathi Advaithi, CEO of mid-cap datacenter company Flex, acknowledged "a lot of noise this week," but said DeepSeek itself doesn't represent anything new in terms of demand for AI infrastructure.

"At the end of the day, compute density is still a big deal," she said. "We think lower cost in applications like DeepSeek is a good thing for the industry as a whole because it's going to drive a stronger growth in terms of the market itself."

In addition, Advaithi said lower barriers to entry could spur more widespread innovation in AI, driving additional demand for infrastructure providers like Flex.

"We haven't seen enough growth from non-Mag Seven companies and we'll start to see a lot more of that," she said. "It actually accelerates the move towards AI."