DeepSeek, a Chinese artificial intelligence firm backed by hedge fund High-Flyer, unveiled its open-source reasoning model R1 in January using less powerful chips and limited funding. The announcement triggered investor panic, causing a steep Nvidia sell-off that slashed CEO Jensen Huang’s net worth before a swift recovery, industry experts say.
In January, DeepSeek, a Chinese artificial intelligence company owned by High-Flyer, an investment group, unveiled R1, an open-source reasoning model that competes with others. According to the company, R1’s huge language model was constructed using less powerful chips and a fraction of the resources allocated to the most popular, Western-made AI models.
The market value of Nvidia dropped by $600 billion as a consequence of investors selling off their shares in response to this news. During the rout, Huang personally saw a temporary decline of approximately 20% in his net worth. Since then, the stock’s value has largely rebounded.
In Thursday’s pre-recorded interview(via Business Insider), which was produced by Nvidia partner DDN and was part of an event introducing DDN’s new software platform, Inifinia, Huang indicated that investors’ misperception was the cause of the significant market response.
If less processing power is needed to train models, investors have questioned whether Big Tech companies really need to spend trillions on AI infrastructure. Jensen stated that post-training methods, which enable AI models to draw conclusions or make predictions after training, continue to necessitate computational power from the industry.
He went on to say that computing power from Nvidia chips will be in high demand as post-training methods expand and diversify.
“From an investor perspective, there was a mental model that the world was pre-training and then inference. And inference was: you ask an AI a question, and you instantly got an answer,” he said at the event on Thursday. “I don’t know whose fault it is, but obviously that paradigm is wrong,”
The most crucial aspect of intelligence is post-training, yet pre-training is still significant. Huang stated that this is the place where one learns to resolve issues.
His words energizing the AI community were DeepSeek’s ideas.
According to Huang, it is really thrilling. Incredibly, R1’s decision to become open-sourced has generated energy all over the globe.
Although Huang did not publicly address the matter until Thursday’s event, Nvidia representatives had already addressed the market reaction with identical written remarks.
For months, Huang has been fighting the mounting worry that model scaling is in danger. In the months leading up to DeepSeek’s January debut, rumors circulated that OpenAI’s model advancements were stalling, casting doubt on the veracity of the AI boom and, by extension, Nvidia’s ability to maintain its current rate of profit.
According to Huang’s November remarks, scaling is still very much in use; the focus has changed from training to inference. On Thursday, Huang added that the post-training approaches are “really quite intense,” and that models would continue to get better with the addition of additional reasoning processes.
If Huang’s DeepSeek remarks are any indication, Nvidia’s first earnings call of 2025, set on February 26, will be a sneak peek. From Palantir to Airbnb, DeepSeek has been a common subject of conversation during earnings calls for firms in the digital industry.
The subject was posed to AMD earlier this month, a competitor of Nvidia. AMD CEO Lisa Su stated that DeepSeek is leading innovation that is “good for AI adoption.”
Luis Gochoco is a seasoned managing editor and writer with over a decade of experience covering politics, technology, gaming, and entertainment news. With a keen eye for breaking stories and in-depth analysis, he has established himself as a trusted voice in digital journalism. Luis is one of the key forces behind the success of GameNGuide, contributing to 12 million views through engaging and high-traffic content. He also played a pivotal role in generating 8 million views on International Business Times, shaping the platform’s technology and gaming coverage.
Discover more from Today's Esquire
Subscribe to get the latest posts sent to your email.