Nvidia Stock Had A Bad Week After DeepSeek Unveiled Its New Model In January—Now CEO Jensen Huang Says Investors Are Misreading AI's Future Impact: 'Obviously, That Paradigm Is Wrong'

Jensen Huang, CEO of Nvidia Corporation NVDA, addressed investor concerns regarding the necessity of massive AI infrastructure spending during a virtual event uploaded on Thursday.

What Happened: Huang addressed the DeepSeek-led sell-off of the Nvidia shares and said, that investors misinterpreted the DeepSeek’s AI advancements. The pre-recorded event, produced by Nvidia partner DDN, was part of the launch of DDN’s new software platform, Inifinia.

The major reason for the sell-off was due to investors questioning the need for massive AI infrastructure expenditure by Big Tech firms, particularly if models can be trained with lesser computing power. This is a result of DeepSeek’s advancements in AI, which investors may have misunderstood, triggering a significant market reaction.

The Nvidia Chief clarified that the industry still demands significant computing power for post-training methods, which enable AI models to generate predictions or reach conclusions after training. He emphasized that as these methods diversify and expand, the demand for Nvidia’s computing power will also increase

“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,” Huang said. He added, “I don’t know whose fault it is, but obviously that paradigm is wrong.”

About the Liang Wegfeng-led Chinese AI firm, Huang said, “It is so incredibly exciting. The energy around the world as a result of R1 becoming open-sourced — incredible.”

SEE ALSO: NVIDIA’s New AI System Aims To Transform Genetic Research With Unmatched Speed And Precision

Why It Matters: Huang’s comments are critical as it comes ahead of Nvidia’s earnings results scheduled on Feb 26. This is the first time that the CEO has spoken about the DeepSeek-led sell-off and addressed investor concerns. He also openly shared his views on DeepSeek’s R1 model and the path ahead for the fast-evolving AI landscape.

Also in the past, Huan had stressed that model scaling would stay relevant. He reiterated the same in the virtual meeting and said that post-training methods are “really quite intense,” and that models would continue to improve alongside the evolution of reasoning methods.

After the launch of DeepSeek’s R1 model, investors sold off Nvidia stock, leading to a $600 billion drop in market capitalization. As a result, Huang temporarily saw nearly 20% of his net worth erased. However, the stock has since regained much of its lost value.

Meanwhile, renowned investor Michael Del Monte remains optimistic about Nvidia’s growth trajectory. He highlighted the industry’s focus on model optimization and the need for more AI data centers to meet enterprise demand.

Del Monte predicts that AI capital expenditures will increase by nearly 40% year-over-year to reach $313 billion this year, with growth expected to stabilize at these levels in subsequent years.

Meanwhile, Nvidia continues to innovate, recently offering its RTX 5090, 5080 Founders Edition GPUs to loyal users to combat scalpers.

• READ MORE: Elon Musk’s X Hit With $1.42 Million Fine From Brazil Judge He Once Labeled ‘Dictator’

Image via Shutterstock

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.