Jen-Hsun Huang, Mistral AI, and A16z: Building National Sovereignty AI to Prevent Digital Colonialism Starting from Ourselves

NVIDIA CEO Jen-Hsun Huang (Jensen Huang), Mistral AI founder Arthur Mensch, and well-known venture capital firm a16z discussed the necessity of Sovereign AI ( on March 20. The three emphasized that every country should own, train, and operate its own AI, rather than relying entirely on technology from Silicon Valley or China. They further highlighted that AI is a reflection of culture, carrying with it language and values. AI is the power revolution of this generation; if countries around the world do not create their own AI models, they will inevitably become "digital colonies" of other nations.

AI is a general technology that comprehensively impacts healthcare, education, agriculture, and national defense.

Jen-Hsun Huang and Mensch both pointed out that AI has already become a General Purpose Technology ), much like electricity and the internet in the past, penetrating every industry and social aspect, including healthcare, education, agriculture, and even national defense.

The two stated that AI can perform tasks and make judgments like humans. However, this also means that if a company monopolizes this technology, other countries and companies may be forced to "wait and see" as the other party decides the future.

AI is culture, carrying language and values.

AI is not just algorithms and data, but a reflection of "culture." Mensch reminds us that if countries around the world do not participate in training AI in their own languages and cultures, they will be dominated by "foreign cultural AI" unconsciously.

Jen-Hsun Huang added that the Swedish people would not want an external company to decide how the Swedish language should be expressed. Similarly, Middle Eastern countries also need their own linguistic and regulatory understanding. "Your AI should not be decided by others."

Build sovereign AI starting from the infrastructure all the way up to the application layer.

For countries or enterprises, what should be done about Sovereign AI (Sovereign AI)? Here are two suggestions:

Purchase infrastructure: such as chips, cloud resources, and open-source models.

Self-developed upper layer: trained with local knowledge and regulations to create a dedicated language and application system.

Establishing an AI Human Resources Department: In the future, companies will not only manage real human employees but also continuously recruit, train, and evaluate their "AI digital employees."

Future work schematic, AI robot assistants working alongside you. Exclusive models represent national strength; language, medicine, and law should be trained by ourselves.

Mensch believes that model specialization is divided into three layers:

Language proficiency: First understand what you are saying.

Industry specialization: Knowing better in which field you are saying these things.

Local cultural expertise: Finally understanding your values and approach to dealing with things.

He emphasized that this is the value of sovereign AI; no global giant cares more about your own language and culture than you do.

Not coming by oneself is digital colonialism; culture and wisdom cannot be outsourced.

When discussing the most serious risks, the host bluntly pointed out: "If AI is cultural infrastructure and we do not own it, the result will be similar to past colonization."

Jen-Hsun Huang agreed: "Your culture and preferences will directly reflect in AI behavior, this is not something that can be outsourced with an API." He gave an example that while McDonald's and Starbucks can be seen globally, what truly defines a city's culture are the corner coffee shops and local breakfast places.

McDonald's vs Taiwanese Breakfast Shops The Importance of Open Source Models, AI Needs a "Democratic Foundation"

Mistral and NVIDIA are both supporters of open-source models. They believe:

Open source can drive innovation in niche industries (, such as mining, energy, and defense, etc. ).

Open source allows more people to participate, review, and enhance security.

Closing the weight will not enhance national security, it will only allow other countries to take the lead.

From AI to culture and then to economy, the key lies in talent and partnerships.

For the national leaders who want to develop sovereign AI, the two provided the following suggestions:

Strengthening local AI talent cultivation: Without technical personnel, model training and customization are out of the question.

Establish infrastructure and software platforms: including AI factories, data centers, model training environments, etc.

Work with trusted technology companies: partners like Huida and Mistral that provide common tools and support open source.

Make good use of AI to bridge the technological gap: it is not about causing unemployment, but about enabling more people to use AI to solve problems.

The next wave of AI will move from chatting to thinking, from the screen to the physical world.

In the next ten years, Jen-Hsun Huang predicts that AI development will evolve in three directions:

Agentic AI (: Able to proactively handle tasks.

Physics AI ): capable of understanding physics, chemistry, and manufacturing principles.

Robot AI (Physical AI): Truly entering the physical world to assist in production, manufacturing, and healthcare.

This conversation repeatedly emphasized one thing: "No global company cares more about its own language and culture than you do; you cannot wait for others to help with this matter, you must start with yourself."

(Sam Altman’s three observations: Costs dropping by 10 times every year, AI agents becoming the new work standard, and assets that cannot be replaced by AI will appreciate in value)

This article Jen-Hsun Huang, Mistral AI, and a16z: Building National Sovereignty AI to Avoid Digital Colonialism Starts with Ourselves first appeared on Chain News ABMedia.