Woofun AI reports that NVIDIA, the world's largest company by market capitalization, convened its 2026 annual general meeting of shareholders on June 24 local time, where founder and CEO Jensen Huang delivered a definitive assessment of the artificial intelligence sector's trajectory. Huang stated unequivocally that the question regarding return on investment in AI already possesses a clear answer, framing the current technological shift not as a temporary trend but as a foundational era lasting several decades. He characterized modern AI data centers as factories dedicated to creating tokens, which serve as the fundamental units of profit that can be converted into code, answers, designs, actions, and services. While acknowledging that NVIDIA's systems may not represent the lowest initial purchase price, Huang argued they deliver the lowest cost per token generated while achieving the highest throughput, thereby maximizing revenue potential for clients who are effectively investing in income-generating AI factories rather than merely acquiring server hardware. This strategic logic directly underpins the company's exceptional financial performance recorded in 2025, a year where annual revenue climbed 65% to reach $216 billion. Operating income surged 60% to $130 billion, while diluted earnings per share jumped 67% to $4.90, supported by an operating cash flow of $103 billion and a total shareholder return of $41 billion. The data center business segment specifically drove this expansion, posting a 68% revenue increase to $194 billion.

Huang revealed a committed capital allocation strategy based on confidence in sustainable market growth and robust free cash flow generation, pledging to return 50% or more of free cash flow to shareholders annually for the current year, next year, and beyond. The company intends to progressively increase stock buyback efforts and dividend payments over time, signaling a long-term commitment to shareholder value amidst the infrastructure boom. Huang described 2026 as an extraordinary year for both NVIDIA and the broader computing industry, emphasizing that the company is constructing the essential computing infrastructure for a new era where AI represents a fundamental transformation rather than just a software model. For six decades, computing has focused primarily on retrieving, storing, and transmitting information, but AI is now reinventing the field to enable the creation of intelligent systems. Huang outlined a historical pattern where the computer industry undergoes major transformations every 10 to 15 years, tracing the evolution from mainframes to personal computers, from personal computers to the Internet, from the Internet to the cloud, and subsequently to mobile clouds. He asserted that the current transition is even more significant because AI enables computers to understand, reason, plan, use tools, and perform useful tasks, transforming them from passive tools into active assistants capable of utilizing tools themselves. Consequently, data centers have evolved from simple toolboxes into AI factories composed of digital assistants that serve as the infrastructure for producing digital intelligence.

The evolution of generative AI has accelerated rapidly over the last two years, moving from the initial capabilities of ChatGPT to write, draw, summarize, and answer questions, to reasoning-based AI that learned how to think, and now to the arrival of agent AI. These agents possess the ability to use tools, access memory, write code, call other agents, test results, and continue working autonomously until a task is completed. Huang projected that this phase of development will span several decades and be comparable to the construction of essential infrastructure such as power grids, transportation systems, and the Internet, describing it as the largest infrastructure project in human history. In this new paradigm, tokens act as the basic units of intelligence created within AI factories, generating revenue through commercialization where higher hash rates produce more tokens and consequently higher revenue. Driving the next phase of growth, the Vera Rubin architecture has now entered full production, while the new generation of Blackwell AI chips has significantly expanded NVIDIA's infrastructure coverage across diverse customer segments including hyperscalers, cloud service providers, AI research institutions, industrial enterprises, and sovereign entities. Model developers and hyperscalers have already deployed hundreds of thousands of Blackwell GPUs, and the construction of AI factories is expanding rapidly across key industries with companies such as Capital One, Hyundai Motor Group, Jane Street, and Eli Lilly adopting NVIDIA's infrastructure to deploy AI solutions.

Per Woofun AI, the company's international revenue has seen a threefold increase, exceeding $30 billion, driven by nearly 40 countries and regions with a combined GDP of $50 trillion building AI factories powered by NVIDIA's infrastructure. Huang emphasized that AI infrastructure is no longer experimental but has entered the production phase, constituting a new industry that can be visualized as a five-layer cake comprising energy, chips and systems, infrastructure, models, and applications. Unlike traditional data centers that store and provide file services, AI factories create tokens that transform into tangible outputs like code and services. Thanks to its end-to-end, fully integrated infrastructure and a large ecosystem of partners, NVIDIA drives the construction of these AI factories in unique ways, with agent AI accelerating infrastructure investments because this marks the first time AI is being used to perform practical tasks and create real economic value. As more organizations seek to mass-produce intelligence, the demand for AI factories will far exceed what cloud computing can meet today, extending to enterprises, sovereign nations, and regional AI clouds. Looking toward the future, Huang identified physical AI as the next major stage of growth, requiring another round of infrastructure investment where NVIDIA offers AI infrastructure, Omniverse simulation and digital twin platforms, open-source models, the Jetson Thor embedded computing platform, and software stacks for large-scale development and deployment. Over time, AI will move from the digital world into robot taxis, humanoid robots, and industrial systems, enabling these machines to perceive, reason, and act autonomously in the physical world. As of the close of trading on the US stock market on June 24 local time, NVIDIA's shares were trading at $199 per share, a slight decrease of 0.52%, with a total market cap of $4.82 trillion. This marks a pivotal moment where the theoretical potential of AI has solidified into a tangible, multi-decade industrial reality.