One CEO seems to be unphased by the startup’s emergence. Microsoft chief executive Satya Nadella asserted DeepSeek’s David to the established AI sector’s Goliath could actually be good news for the tech industry as a whole.

“Jevons paradox strikes again!” Nadella wrote on LinkedIn Monday, referring to a theory that increased efficiency in a product’s production drives increased demand. “As AI gets more efficient and accessible, we will see its use skyrocket, turning it into a commodity we just can’t get enough of.”

Last week, emerging Chinese AI company DeepSeek unveiled R1, an open-source model promising to be on-par with Open AI’s o1 model released last year, which is capable of reasoning and solving math problems. The release turned heads. Not only was DeepSeek a potential burgeoning rival to OpenAI, but it also reportedly developed its V3 large language model on a $5.6 million budget—a fraction of the $100 million OpenAI poured into training ChatGPT-4. DeepSeek also claims it requires fewer computational resources than competitors.  

The effectiveness of DeepSeek—combined with its reported shoestring budget—has made it a threat to the U.S.’s stronghold over AI development. The U.S. has imposed export controls and restrictions on sales of advanced semiconductors to Chinese companies since 2022 to address national security, which has hampered Chinese competitors. Venture capitalist Marc Andreessen stirred Cold War sentiments and called DeepSeek’s model unveiling “AI’s Sputnik moment,” referring to the day in 1957 when the Soviet Union beat the U.S. in launching the first satellite into orbit.

DeepSeek did not respond to Fortune’s requests for comment.

The markets reacted accordingly: Nasdaq 100 futures plummeted nearly 2.5% Monday, and the S&P 500 slid 1.8%. Microsoft shares tumbled 3.8%, and Meta and Alphabet weren’t far behind, falling 3.1% and 3.3%, respectively. AI darling Nvidia took the biggest beating, with shares falling 13% in premarket trading.

What is Jevons paradox?

But rather than see DeepSeek as a threat to the U.S. tech industry as a whole, Nadella instead saw its barnstorming reveal as a positive in Big Tech’s race to expand AI into the crevices of human life. 

At the foundation of Nadella’s optimism is a 160-year-old theory called Jevons paradox. English economist William Stanley Jevons originated the theory in 1865 after observing a massive increase in England’s coal consumption after the invention of the Watt steam engine, which required fewer resources than previous steam engine models. But rather than resulting in less coal being used, the widespread use of the more-efficient Watt engine resulted in more coal consumption. By citing Jevons paradox, Nadella suggests more efficient AI, as exemplified with DeepSeek’s R1 model, will increase the demand for the technology.

There’s a hangup with Nadella’s supposition: Jevons initially posited that increased efficiency of a resource would paradoxically lead to that resource diminishing at a faster rate than if its use were less efficient. For resource-guzzling technology like AI, increased demand could come at the expense of the environment. Economic historian Albert Schram argued, using the framework of Jevons paradox, more efficient AI would result in the use of more data storage and natural resources. The environmental impact of this AI surge could make its continued development economically unsustainable in the long run. 

“Ultimately, the application of Jevons Paradox to AI highlights the need for careful consideration of the potential unintended consequences of technological advancements and the importance of taking a proactive approach to address these issues,” Schram said in a May 2023 LinkedIn post.

Regardless if Jevons paradox comes to fruition, Nadella understands the start-up isn’t to be trifled with. Last week, at the World Economic Forum in Davos, Switzerland, the Microsoft boss did not shy away from acknowledging the technology as a real contender in the world of AI.

“To see the DeepSeek new model, it’s super impressive in terms of both how they have really effectively done an open-source model that does this inference-time compute, and is super-compute efficient,” Nadella said Wednesday. “We should take the developments out of China very, very seriously.”