Jensen Huang of Nvidia Picks 'Physics' as the Hottest Major... Why? [Tech Talk]

Spotlight on Unpopular Pure Sciences Attracts Tech CEOs
Jensen Huang: "If I Could Go Back, I Would Study Physical Sciences"
The Next Stage of AI Is Understanding Physical Interaction

Jensen Huang, CEO of Nvidia, recently stated in an interview with CNBC in the United States, "If I could go back to my twenties, I would study physical sciences." Physical sciences encompass a broad range of basic sciences, including physics, astronomy, and earth science. Traditionally, as a pure science, it has been considered less helpful for employment. However, as the advancement of artificial intelligence (AI) now threatens computer science, pure sciences are being reevaluated.

Spotlight on Unpopular Pure Sciences Once Deemed Unhelpful for Employment

Jensen Huang, CEO of Nvidia. Photo by AFP Yonhap News

Physical sciences align with the AI vision that CEO Huang envisions. At the 'Hill & Valley Forum' held in Washington D.C. in April, he stated, "After the current era of reasoning AI, 'physical AI' will arrive," describing it as "an AI that understands physical laws, friction, inertia, and causality."

The 'reasoning AI' Huang referred to includes familiar models such as ChatGPT and Gemini. These AIs excel at compressing images or vast amounts of text, and are adept at composing text or creating images and videos.

'Physical AI' goes further by understanding the physical laws themselves and predicting the outcomes of specific physical phenomena. This means it could instantly calculate the force needed to gently grip soft fruit, the fuel efficiency based on a car's weight, or the friction experienced by a spacecraft or aircraft during flight.

Dassault Syst?mes' computational fluid dynamics (CFD) simulation program. Traditional physics simulations required expensive and heavy software and hardware. Dassault Syst?mes website

Until now, these tasks have been extremely complex and time-consuming. They have required dozens of specialized engineers, professional software such as computational fluid dynamics (CFD) programs, and very expensive supercomputers. If AI's intuition can make such simulations faster and more accurate, it could revolutionize the field of physics. Countless theoretical verification tasks that have been delayed due to lack of budget for expensive software and supercomputers could be accomplished in an instant.

Additionally, it would be extremely useful for testing new aerospace designs, analyzing the results of particle accelerators, or building humanoids. It could also accelerate efforts to understand atmospheric interactions in greater detail to address climate change, and to unlock the mysteries of the universe. As a result, the scope of activities for human physicists will expand beyond academia to the broader economy.

Computer Science Declines, Physics Rises... Will AI Determine the Value of Academic Disciplines?

Google DeepMind's AI simulating the joints and tendons of the human leg. The advancement of 'physical AI' requires the assistance of experts with knowledge in physics. YouTube capture

Other big tech founders also agree that the value of pure sciences, including physical sciences, will rise. Last month, Pavel Durov, founder of Telegram, wrote on his X account, "If you are a student, major in mathematics. Mathematics trains your brain in logical thinking and problem solving, which are essential for managing and growing a business." In response, Elon Musk, founder of Tesla, replied, "And you should also learn physics."

The rise of pure sciences coincides with the decline of computer science, which has long been the most popular major. According to the U.S. current affairs magazine The Atlantic, the number of computer science majors in the United States surged by 400% from 2005 to 2023, but last year, the number of freshmen increased by only 0.2%. This year is expected to be the first year in which the number of new computer science students does not grow.

Currently, the occupation being most rapidly replaced by generative AI is programmer. Major AI companies such as OpenAI (ChatGPT), Google (Gemini), and Anthropic (Claude) are releasing the latest coding AIs in quick succession. Now, even people with no prior knowledge of programming can obtain complex source code simply by entering a few basic requests.

As a result, big tech companies are cutting general programmer jobs and reallocating those budgets to AI research. According to Layoffs, a U.S. IT layoff tracking site, nearly 60,000 employees were laid off from U.S. big tech companies in the first half of this year. We have entered an era where the status of jobs and academic disciplines is determined by the advancement of AI.