From the way we work to the way we stay healthy, the AI revolution has the potential to greatly improve every aspect of our lives.

However, to ensure that the United States and other countries can help shape the trajectory of this technology, it cannot rely solely on the technologies developed by private enterprises.

University research has driven the development of artificial intelligence, laying the foundation for the commercial boom in AI that we are experiencing today. Equally important, the leaders of those pioneering AI companies are inseparable from the cultivation of academia.

But now, large foundational models (LFM) such as ChatGPT, Claude, and Gemini require powerful computing power and vast datasets, which private companies can better meet these needs. As a result, they have replaced academia and stood at the forefront of the field of artificial intelligence.

Empowering our universities to stand at the forefront of AI research alongside private enterprises will be key to realizing the long-term potential of the field. Achieving this requires changing the unequal status of access to computing resources between academia and industry.The greatest advantage of the academic community lies in its ability to undertake long-term research projects, as well as fundamental research that breaks the boundaries of knowledge. The freedom to explore and experiment with bold, cutting-edge theories will lead to new discoveries and innovations, laying the foundation for future advancements.

Although tools supported by large foundational models are accessible to everyone, there are many unanswered questions because they remain a "black box" in many aspects.

For instance, we know that artificial intelligence models have a tendency to generate "hallucinations," but we are still unclear about the exact reasons behind this phenomenon.

Universities can chart a future where artificial intelligence truly benefits humanity, as they are not influenced by commercial and market forces. Expanding the academic community's access to resources will make the research and application of artificial intelligence more inclusive.

U.S. President Joe Biden issued an executive order on artificial intelligence in October 2023, authorizing the pilot project of the National Artificial Intelligence Research Resource (NAIRR). This is a step in the right direction.By collaborating with private enterprises, NAIRR will create a shared research infrastructure for artificial intelligence.

 

If it fully realizes its potential, it will become an important core that helps academic researchers to gain GPU computing power more effectively. However, even if NAIRR is well-funded, its use of resources may still be fragmented.

 

As some have suggested, this issue could be mitigated if NAIRR focuses only on a certain number of dispersed projects.

 

But we should also seek more creative solutions to ensure that enough GPU resources are in the hands of the academic community. Here are some of our preliminary ideas.

 

Firstly, we should utilize large-scale GPU clusters to improve and leverage the supercomputing infrastructure in the United States. Academic researchers should be able to collaborate with the U.S. national laboratories to tackle significant challenges in artificial intelligence research.Secondly, the U.S. government should explore ways to reduce the cost of high-end GPUs for academic institutions, such as providing grants or R&D tax credits and other financial assistance.

New York has taken some initiatives in this regard, making universities key partners in the state's development of artificial intelligence and playing an important role. This model should be emulated across the United States.

Lastly, some export control restrictions may lead to an excess inventory of high-end AI chips for U.S. chip manufacturers.

In this case, the government can purchase these chips and distribute them to universities and academic institutions.

These actions will trigger a surge in AI research and innovation in academia. Ambitious researchers in universities have many diverse ideas, but are often held back due to a lack of resources.Providing universities with adequate computing resources will enable their work to complement the research of private enterprises.

 

Thus, the academic community can become an indispensable hub for technological advancement, fostering interdisciplinary collaboration, conducting long-term research, nurturing the next generation of AI talent, and promoting ethical innovation.

 

Historically, similar investments have played a key role in innovation. After World War II, the United States established a symbiotic relationship between the government, academia, and industry, which took humanity to the moon, sowed the seeds of Silicon Valley, and created the internet.

 

We need to ensure that the academic community maintains a strong position in the innovation ecosystem, and investing in its computing capabilities is the necessary first step.