OpenAI, AMD, CoreWeave, and Microsoft gather in Congress: The three major battlefields of AI infrastructure, chips, and education are launching.

The U.S. Senate held an AI Future Policy Hearing this morning on May 9, gathering the four major CEOs from OpenAI, AMD, CoreWeave, and Microsoft. They discussed AI chips, energy infrastructure, open ecosystems, as well as the democratization of AI and talent cultivation. The tech giants clearly expressed one thing: "This is a technological competition concerning national security, the economy, and the future of human civilization. The U.S. cannot stop and cannot lose."

Sam Altman was the first to indicate that the AI revolution is no longer in the future, it is now.

OpenAI CEO Sam Altman stated that ChatGPT has now become the fifth largest website in the world, with weekly active users exceeding 500 million. He emphasized:

Scientists' productivity increases by 2 to 3 times.

The public can obtain better medical advice and education.

The United States must expand more infrastructure, especially in power and computing capacity.

Su Zifeng stated that AI is a global competition, and while the United States is leading, it must not slack off.

AMD CEO Lisa Su stated:

Although the United States currently leads in the AI field, this is a marathon.

It is pointed out that the United States must maintain its lead in chip manufacturing, software integration, and application ecosystems.

She urged the government to do five things:

Accelerate investment and production capacity

Maintain innovation and openness

Strengthen the local supply chain

Invest in AI talent education

Manage export controls well but do not restrict the dissemination of technology.

CoreWeave started from a garage and is now an AI infrastructure giant.

CoreWeave CEO Michael Intrator approached the issue from the perspective of industry scale and energy consumption, stating:

The company grew by 12,000% in just two years, with revenue reaching 1.9 billion USD.

30 data centers with 250,000 GPUs

AI computing power has grown 100,000 times compared to 2018, traditional cloud computing simply cannot support it.

He emphasized that you cannot support the economy of the 21st century with the infrastructure of the 20th century, and that AI data centers require:

super computing power

Ultra-high-speed internet

Super Cooling System

Ultra-high energy supply

He proposed four major policy recommendations:

Stabilizing Policies and Supply Chains: Encouraging Industry to Invest

Fast power generation and grid construction: Energy policy must be reformed.

Balancing export controls with global market expansion

Strengthening public-private cooperation and talent training: Currently collaborating with Microsoft and Princeton University.

Microsoft's Brad Smith: Don't underestimate AI, and don't underestimate humanity.

Microsoft President Brad Smith concluded the event with a single statement:

"We are not building machines that are stronger than humans, but rather machines that help humans become better."

He started by talking about the "AI technology stack diagram," emphasizing:

The leading AI in the United States relies on the overall ecosystem, where every level is indispensable (from power plants to developers).

Three key points: innovation, diffusion, export

He also reminded the U.S. government that the United States has only 4.5% of the global population, and to lead the world, it must serve the world. To gain the world's trust in American technology, it cannot just be about blocking; trust must be earned.

AI is the next national movement for the United States, with speed, talent, and trust being the key factors.

In summary, this hearing is not just a discussion on technology policy, but a national-level AI mobilization order. Four industry leaders conveyed the same message at different levels:

"The AI revolution is already on the way. If the United States wants to win, it cannot be complacent and must involve everyone."

This article discusses OpenAI, AMD, CoreWeave, and Microsoft gathering at Congress: the three major battlefields of AI infrastructure, chips, and education are starting to unfold. Originally appeared in Chain News ABMedia.