How Does AI Drive Global Economic Growth?

TL;DR

AI can drive economic growth by enabling computers to understand unstructured information, reason from context, and perform tasks from natural-language instructions. Its economic impact depends on a five-layer stack of energy, chips, cloud infrastructure, models, and applications, with applications in healthcare, financial services, manufacturing, robotics, transportation, and science creating the practical benefits.

Transcript

Good morning everyone. It's really nice to be back here in the Congress Hall. Hopefully everybody had a good day yesterday and are enjoying it today. It is my real pleasure to introduce Jetson Wong, who is somebody I admire, somebody I've watched, and somebody who has been a teacher to me on the journey of learning about technology and AI. It is am... Read More

Key Insights

• AI is a new computing platform because developers can build applications on top of models such as ChatGPT and Anthropic Claude. Like previous shifts to personal computers, the internet, mobile devices, and cloud computing, it reinvents the computing stack and enables new categories of applications.
• Traditional software is effectively prerecorded because humans explicitly describe algorithms and recipes for computers to execute. It primarily works with structured records and database queries, while AI can interpret unstructured images, text, and sound, understand their meaning, and reason about an appropriate action.
• AI processes context in real time and can infer a user's intention from loosely structured natural-language prompts. When the system understands the requested outcome, it can reason about the available information and perform a task instead of merely retrieving a predefined response from structured data.
• Industrial AI is a five-layer stack consisting of energy, chips and computing infrastructure, cloud services, AI models, and applications. The model layer receives much public attention, but it cannot operate without the physical and digital infrastructure provided by every supporting layer beneath it.
• Applications are the layer where AI's economic benefits are expected to materialize. The transcript identifies financial services, healthcare, manufacturing, robotics, transportation, and science as areas of opportunity, while emphasizing that useful applications depend on models becoming capable enough to support reliable development.
• AI infrastructure is undergoing what Huang calls the largest infrastructure buildout in human history. The expansion includes energy capacity, chip factories, computer factories, cloud systems, memory production, and AI factories that process context so models can generate intelligence for applications.
• AI models became better grounded and more capable of research and reasoning during the previous year. Huang says they can analyze circumstances that may not have appeared in training, divide problems into step-by-step reasoning tasks, create plans, answer questions, and perform requested work.
• AI-native companies attracted most venture funding during 2025, according to the discussion. These companies operate across healthcare, robotics, manufacturing, financial services, and other major industries, reflecting increased confidence that improved models are now capable enough to support commercially useful applications.
• More videos from this channel:
  • How Will AI and Tokenization Drive Crypto Growth?
  • Why Is Nvidia Betting on Agentic AI Factories?
  • How will Airbnb become a bigger platform with AI
• More from this channel: Yahoo Finance

Questions & Answers

Q: How can AI drive global economic growth?

AI can drive global economic growth by supporting useful applications across healthcare, financial services, manufacturing, robotics, transportation, and science. It also stimulates investment throughout its supporting stack, including energy, semiconductor plants, computer production, cloud services, memory, and AI factories. The broadest benefits emerge when capable models are converted into applications that improve tasks, services, and productivity across many sectors.

Q: What makes AI different from traditional software?

Traditional software is effectively prerecorded because developers specify the algorithm or recipe that a computer must execute. It generally retrieves and processes structured information stored in defined fields and tables. AI can instead interpret unstructured text, images, and sound, consider contextual information, infer an intention expressed through a natural-language prompt, reason about that information, and perform a suitable task in real time.

Q: What are the five layers of the industrial AI stack?

The five layers are energy, chips and computing infrastructure, cloud infrastructure and services, AI models, and applications. Energy supports the real-time processing required to generate intelligence. Chips provide computing capacity, cloud systems deliver infrastructure and services, models interpret and reason about information, and applications use those capabilities in fields such as healthcare, financial services, manufacturing, and robotics.

Q: Why does AI require so much infrastructure?

AI requires substantial infrastructure because it processes context and generates intelligence in real time. The applications at the top of the stack depend on models, cloud services, computing systems, chips, and energy beneath them. Expanding AI therefore requires capacity across the entire chain, including chip fabrication plants, computer factories, memory production, cloud infrastructure, and AI factories that process information for models and applications.

Q: Where will the economic benefits of AI appear?

The economic benefits are expected to appear primarily in the application layer rather than in models alone. Applications can use AI capabilities within financial services, healthcare, manufacturing, robotics, transportation, science, and other major industries. These applications turn a model's ability to understand unstructured information, research questions, reason through circumstances, and perform tasks into practical services and productivity improvements.

Q: What is agentic AI according to Jensen Huang?

Agentic AI describes the evolution of language models into systems that can do more than generate interesting responses. Huang says improved systems are better grounded, can conduct research, and can reason about circumstances they may not have encountered during training. They can break a problem into step-by-step reasoning stages, develop a plan, answer a question, or carry out a requested task.

Q: Why are open AI models important for broader adoption?

Open models were identified as one of the major breakthroughs in the AI model layer. The discussion presents DeepSeek as a major event for industries and companies around the world, although the supplied transcript ends before the full explanation is completed. Within the available remarks, open models are positioned as an important development alongside better grounding, research capabilities, reasoning, and agentic systems.

Q: Why are investors funding AI-native companies?

Investors are funding AI-native companies because the models have become capable enough to support applications built on top of them. The discussion says 2025 was one of the largest years for venture funding and that most funding went to AI-native companies. Those businesses span healthcare, robotics, manufacturing, financial services, and other large industries where AI applications may create economic value.

Summary & Key Takeaways

• AI represents a computing platform shift comparable to earlier transitions involving personal computers, the internet, mobile technology, and cloud computing. Unlike traditional software built from prerecorded instructions and structured database queries, AI can interpret images, text, sound, context, and loosely expressed intentions, then reason about the appropriate response or task.

• The industrial AI stack consists of five connected layers: energy, chips and computing infrastructure, cloud services, AI models, and applications. Every application depends on the supporting layers below it. Because intelligence is generated through real-time processing, expanding AI use requires substantial investment in energy systems, semiconductor fabrication, computer production, cloud capacity, and AI factories.

• Improved models are making application development more practical across healthcare, financial services, manufacturing, robotics, transportation, and science. Huang says models became better grounded, more capable of research and step-by-step reasoning, and increasingly useful as agentic systems. Open models also broaden access, while investment in AI-native companies signals growing commercial activity above the model layer.