Sequoia Dialogue with Jensen Huang: Computing Model Undergoes a 60-Year Transformation; You Won't Be Replaced by AI, But You Will Be Dimensionality-Reduced by 'Those Who Master AI'

marsbitPublished on 2026-06-12Last updated on 2026-06-12

Abstract

NVIDIA founder and CEO Jensen Huang, in a conversation with Sequoia Capital's Konstantine Buhler, argues that we are witnessing the most significant computing shift in 60 years—from retrieval-based to generative computing. Instead of just storing and retrieving data, future systems will generate highly personalized content (text, images, video) on demand, powered by massive "AI factories." Huang envisions a global "intelligence network" that will envelop the planet, following the historical patterns of energy and communication grids. He outlines a five-layer investment framework: 1) Energy, 2) Chips/Computers, 3) Infrastructure (data centers), 4) AI Models, and 5) Applications. He predicts this ecosystem will reach a scale of $20 trillion annually. Crucially, Huang pushes back against fears of AI-driven job loss. He distinguishes between specific "tasks" (e.g., typing, analyzing images) and overall "jobs" (e.g., CEO, radiologist). While AI automates tasks, it increases efficiency and demand for the higher-value problem-solving aspects of professions, thus creating more jobs and "up-leveling" careers. The real risk, he asserts, is not being replaced by AI, but being outperformed by someone who effectively leverages it. He urges everyone to embrace AI as a tool for augmented capability and innovation.

Source: Sequoia Capital

Compiled by: Yuliya, PANews

Editor's Note: In the past, our data centers merely stored files for human retrieval; now, computing is shifting towards generation, where every word, image, and video is produced in real-time and highly customized according to the requester's context. In this global wave of transformation, Sequoia Capital Partner Konstantin Buhler and NVIDIA founder and CEO Jensen Huang engaged in an in-depth dialogue, discussing the major shifts in computing technology. Huang believes that automation will not lead to unemployment but rather a comprehensive upgrade in labor demand and a dimensional elevation of professions themselves; people will not lose their jobs because of AI, but they may be replaced by those who are adept at leveraging AI.

AI Factories and the Generational Leap in Computing Models: From Retrieval to Generation

Konstantin: Thank you very much for being here, Jensen. We are in the midst of a massive AI revolution, whose scale and speed may even surpass the Industrial Revolution. You have stated that what is happening now is the largest infrastructure construction in human history. At the heart of this construction are AI factories, and the company powering all of this is NVIDIA. Can you tell us what an AI factory is and why it is the most worthwhile investment for all enterprises in the next decade?

Jensen Huang: There are many ways to understand AI. What the public is most familiar with might be interacting with a chatbot through a web browser: you give it a prompt, and it replies with a passage. Even if you've been using AI for a while, you'll find its capabilities have evolved remarkably over the past two to three years.

Two years ago, everyone heard about ChatGPT. It is essentially computer software that understands the information you input. It can perceive, comprehend information, and transform and generate it into other content. For example, you can give it a PDF file and ask it to summarize it—that's text-to-text. You can also have it generate an image based on a story—that's text-to-image. Or give it a photo and ask it to describe the scene—that's image-to-text. This capability was called generative AI two years ago.

But beyond understanding and generation, thinking ability is even more valuable. The foundation of generative AI endows it with the capacity for internal thought, step-by-step reasoning, and problem-solving. Moreover, it can now generate control instructions to use tools—whether digital tools like browsers, spreadsheets, Photoshop, AutoCAD, or, in the future, to control mechanical systems (which is robotics and autonomous driving).

Two years ago, people found ChatGPT interesting; it could write poems and songs but occasionally talked nonsense. Today, two years later, we have agentic systems. AI is no longer just about understanding information; it can now reason and perform useful work. Because it can do useful work, AI has created real commercial value. We don't pay for friends who only talk big, but we pay people who actually get work done. Now, people are hiring AI by the hour every day, for example, paying it $20 to $30 per hour. That's also why it has become the fastest-growing software business in human history.

Looking upstream from an industrial logic perspective, we must return to first principles. The fundamental concept of the computer industry as we know it today was established roughly 64 years ago. At that time, IBM launched the System/360, which was also the reason IBM became the world's most valuable company then.

For the past 60 years, the essence of computing has been pre-recording and retrieval: you write a story, take a photo, record a video, save it as a file into a hard drive; when you want to use it, you retrieve it from the hard drive. That's why those buildings are called data centers. They just store data and don't do much computing.

But now things have changed. In the AI era, every time you provide new context and a new request, AI performs real-time understanding, reasoning, and generates brand-new results. For instance, the speech I'm giving now is generated in real-time based on the different backgrounds of everyone present, rather than being read from a script. This is what intelligence is.

In the future, every pixel, every sound, every piece of video, even every ad and news article will be custom-made, fully generated for you, rather than pre-recorded and retrieved. This means, we will need a massive number of generators in the future, which are the large-scale computers we are building—these are AI factories.

The Intelligent Network Enveloping the Earth and the Dynamo of the Digital Age

Konstantin: How large will this generator be?

Jensen Huang: Currently, we provide information and intelligence generation for roughly 1 billion people globally. But because AI has become agents that can work on their own, one agent can even communicate and collaborate with another. Within NVIDIA, there might be hundreds or thousands of agents talking to each other and solving problems (of course, they operate within safe sandboxes and guardrails).

This means that in the future, not only will humans be using the internet, but there might also be hundreds of billions of agents working tirelessly day and night on the internet. Agents for enterprises, autonomous vehicles, robots, even systems in every building will all be talking to each other. All instructions, all thinking will be generated in real-time.

It's like a thick computational network, wrapping the entire Earth like a cocoon. This sounds exaggerated, but it has actually happened twice in history:

  • The first time was 300 years ago when Germany's Siemens made a machine. You ignite it, and it outputs a powerful invisible force—electricity. Today, the power generation network (the electrical grid) envelops the entire Earth.

  • The second time was 35 years ago with the birth of the internet in the US, which now also envelops global communications.

Now, we are welcoming the third network after energy and communication: the intelligent network. The business NVIDIA relies on for survival today is building this new era's dynamo. The dynamo 300 years ago input the physical motion of water flow, wind, or coal (atoms) and output electrons; our NVIDIA machine, on the other hand, inputs electrons (electrical energy) and outputs digits. These digits, through different combinations, become language, mathematics, or the language of proteins and human biology, the language of physical laws and climate prediction, even the language of 3D worlds, robotics, and autonomous driving.

These two machines, separated by 300 years, share a similar principle: atoms in, electrons out; electrons in, digits out. These digits are what we call Tokens, which is intelligence. We mass-produce these intelligent Tokens in factories; that's the essence of AI factories.

Konstantin: We are in the midst of a wave where multiple revolutions converge. From the energy transition, routers of global telecommunication networks, to today's GPUs and AI factories at the core of the intelligence revolution, like the H100 or the latest Vera Rubin architecture. Integrating everything needed.

Jensen Huang: Yes, our compute unit is called a "rack." A rack contains 72 chips. This year, we will produce about 8 million of these components. A rack weighs 2 tons, costs $4 million, and has 1.5 million parts. It's the most expensive piece of equipment in the world, but we're mass-producing them like smartphones, shipping them to data centers worldwide. This thing is huge; moving them is definitely heavy labor.

The Five-Layer Cake Investment Logic for Participating in the AI Era

Konstantin: This is a very exciting vision. How can both large enterprises and individuals participate in this revolution?

Jensen Huang: To invest in the AI industry, you can imagine its industrial layout as a five-layer cake. You know, a $50 billion AI factory can generate $300 to $400 billion worth of intelligence; its return on investment is quite astonishing. So, what are these five layers?

The first layer is Energy: That is, the power generators at the very bottom. This is the biggest growth opportunity for the energy industry in generations. To support computing, sustainable energy (nuclear, wind, solar, hydrogen, etc.) will receive massive investments. As long as you can generate energy, you will get investment. That's why companies like Siemens, Mitsubishi, GE Vernova are performing so well now.

The second layer is Chips/Computers: Including chips, computers, networking equipment, switches, and silicon photonics technology.

The third layer is Infrastructure: Including land, power, building shells, capital, and the day-to-day operations of data centers. These resources are currently in extreme shortage.

The fourth layer is the Model Layer: That is, the large models built on cloud infrastructure. This is the most market-driven, capital-intensive field in human history. Well-known examples include OpenAI and Anthropic. But remember, AI can learn not just natural language; it can learn anything with structure. We are learning the laws of the physical world—for example, when I sat down just now, I was very confident not because I had a 47% chance of falling through the chair, but 100% confident in the laws of physics. AI can similarly learn the meaning of proteins, the significance of genes, the function of cells. The industry scale of the physical world is $80 trillion, a crucial frontier that is currently less discussed.

The fifth layer is the Application Layer: Based on the underlying technology, countless startups are reshaping industries like financial services, law, accounting, transportation, and logistics. Last year, venture capital invested $100 billion in this top layer, the highest VC investment year in human history.

This future will be immense. We are merely at the starting line, with an estimated $1 trillion being invested into this ecosystem this year. But I estimate that AI will be a massive ecosystem worth around $20 trillion annually in the future. How important is intelligence? Who needs it? How much do you need? Figure these out, and you'll know the direction for investment.

AI Isn't Here to Take Your Job; It's Here to Help You Level Up

Konstantin: This is not only a market opportunity worth trillions of dollars but also means the explosion in hardware infrastructure and application layers will create a vast number of real jobs for humanity.

Jensen Huang: Absolutely correct, and this point we must emphasize. Right now, attitudes toward AI vary across countries and cultures. But I sincerely want to advise everyone: Beware of those Hollywood sci-fi movie plots. Stop listening to people saying things like "The Terminator is coming," "the technological singularity is here," "there's a 20% chance AI will destroy humanity." That's complete nonsense.

Some even scare others by saying, "We don't even know how AI works; it's too mysterious; maybe it will just walk away tomorrow." That's even more baseless. AI is computers and software; engineers certainly know how it works, otherwise how could they make it safer and smarter every year?

Today's AI has significantly reduced hallucinations; the knowledge it generates is accurate and contextually relevant. When it doesn't know something, it even looks it up. It might even self-reflect before answering you, compare several options, and then tell you the answer. Just as cars today are much safer than they were 100 years ago, the tech world is going all out to make AI extremely safe.

So, focus your attention on what is certain. I am very certain about one thing: You probably won't lose your job because of AI, but you will definitely lose your job to the person who uses AI.

Since this is a technology that can give humans superpowers, you should hurry up and use it! Whether you're telling your loved ones, your children, your company, or your country: you must embrace AI.

Konstantin: But when it comes to jobs, people are genuinely anxious.

Jensen Huang: I get upset every time I hear people creating panic about jobs. This year, we invested $1 trillion into this ecosystem—energy, chips, infrastructure, model layer, application layer—all creating far more jobs than ever before.

Some might ask, what about traditional positions? There's a common cognitive mistake people make here: they confuse "Job" with "Task."

Take me, for example. I'm a CEO. My daily "tasks" mainly involve typing and speaking. Now, AI is far better at typing and speaking than I am—it's superhuman level. But as a CEO, I'm actually busier now than before.

Let me give you a more profound example. About 12 years ago, a top computer scientist stood up and warned everyone, saying computer vision could read medical images tirelessly, never missing a detail, already at superhuman levels. He asserted that the first profession to be eliminated by AI would be "radiologists," advising people not to study this field anymore.

He was completely correct in his technical judgment. Now, all radiology systems integrate computer vision; all radiologists use AI to assist their work. But what was the outcome? The global demand for radiologists actually increased!

Why? Because the purpose of a radiologist is not to read images but to diagnose diseases alongside clinicians. Due to automation, their efficiency greatly increased; hospitals could take in more patients waiting in line, and radiology departments became more profitable. Hospitals found profits rising and patient numbers growing, so they hired even more radiologists! Those who heeded the warning and didn't study radiology missed the opportunity.

Similarly, recently, some said that because AI can write code, 90% of software programming is gone; we no longer need software engineers. But the fact is, we are hiring more software engineers now than ever before! Because the purpose of a software engineer is to solve problems and innovate, not to compete in typing speed. Writing code is just a task; solving problems is the core.

AI will not only not eliminate jobs; it will actually enhance the value of your work. If I were a plumber today, I might just work according to blueprints; but tomorrow, with AI's support, I might also be a kitchen designer. If I were a furniture seller or carpenter, in the past, you only expected me to nail wood together, but with AI, I can directly provide you with full interior design plans, making your home incredibly beautiful. My professional skills have been elevated to a higher dimension!

So I believe that the current narrative about AI causing human unemployment is completely wrong; it's just to scare others away so that they can profit from it. Looking at my entire career, computer technology has become increasingly complex. In the past, people who mastered the C++ programming language only accounted for 2% of the population (maybe more among you in the Silicon Valley venture capital circle). But now, because of AI, as long as you understand human language, you can program. For the first time, we are truly closing the technological gap; we must bring everyone along into this new era.

Related Questions

QWhat is the fundamental shift in computing paradigm that Jensen Huang describes, and how does it differ from the past 60 years?

AJensen Huang describes a shift from a 'retrieval-based' computing paradigm to a 'generative' one. For the past ~60 years, computing was about pre-recording information (like files, photos) and retrieving it when needed. Now, in the AI era, computing is about real-time understanding, reasoning, and generating entirely new content (pixels, sound, video) tailored to specific contexts and requests for each user.

QAccording to the conversation, what is an 'AI factory' and why is it a critical investment?

AAn 'AI factory' is a large-scale computing facility designed for massive generation. It produces 'tokens' of intelligence in real-time. It's a critical investment because the future demands a vast amount of generated, personalized content for humans and billions of AI agents, representing a foundational shift in how computing infrastructure is built and used.

QHow does Jensen Huang frame the five-layer 'cake' of the AI industrial landscape for investment?

AHe frames it as a five-layer structure: 1) Energy (the foundational power source), 2) Chips/Computers (hardware like GPUs, networking), 3) Infrastructure (land, power, data center operations), 4) Models (large language and domain-specific models), and 5) Applications (startups reshaping industries like finance, law, logistics).

QWhat is Jensen Huang's core argument against the fear that AI will lead to massive job losses?

AHis core argument is that people confuse 'jobs' with 'tasks'. AI automates specific tasks (e.g., writing code, analyzing X-rays), but this increases efficiency and demand for the core purpose of the job (e.g., problem-solving, diagnosis). This leads to more job creation and 'upskilling,' allowing professionals (like plumbers or carpenters) to offer higher-value services (like design). He asserts you won't lose your job to AI, but to someone who uses AI effectively.

QWhat historical analogy does Huang use to describe the emerging 'intelligence network' powered by AI?

AHe uses the analogy of two previous global networks: 1) The power grid (emerging ~300 years ago with the dynamo, converting physical motion to electricity), and 2) The internet (emerging ~35 years ago for communication). He states we are now building the third such network: a global 'intelligence network' that generates smart tokens, comparing Nvidia's role to that of building the new 'dynamo' for this era.

Related Reads

The Revelation from the Raydium Theft Incident: New DeFi Vulnerabilities Lurking in Forgotten Old Contracts

**Raydium Exploit Reveals DeFi's Hidden Risk: Forgotten "Zombie" Contracts** A recent attack on Raydium's deprecated V3 AMM pools resulted in a loss of approximately $1.34 million. The hacker exploited pools that were no longer supported by Raydium's current UI or SDK but remained fully functional and accessible on-chain. This incident highlights a critical, often overlooked category of risk in DeFi: inactive or legacy smart contracts that projects fail to properly decommission. Since March 2025, there have been at least 8 publicly reported attacks targeting such abandoned contracts, with total losses around $10.8 million. Including older pools and deprecated features, the count rises to 10 incidents with roughly $22.5 million in losses. These "zombie contracts" represent a lifecycle management failure rather than a code vulnerability, yet they are typically misclassified under general "code bug" categories in security reports, masking the true scale of the problem. The root cause is that projects often merely document a contract as "deprecated" without taking essential technical steps to secure it: withdrawing remaining assets, disabling external call functions, and implementing ongoing monitoring. These forgotten, under-monitored components become prime targets for attackers. To address this, the industry needs to recognize "zombie contracts" as a distinct risk category and establish standardized decommissioning protocols. Essential steps should include: 1) a formal retirement announcement, 2) removal of all front-end integrations, 3) withdrawal of locked assets, 4) disabling key contract functions, 5) ongoing security monitoring, 6) clear user communication, and 7) a post-mortem analysis. The value of a DeFi project lies not only in its current TVL but also in the security of its historical codebase, which has now become a new attack surface.

Foresight News16m ago

The Revelation from the Raydium Theft Incident: New DeFi Vulnerabilities Lurking in Forgotten Old Contracts

Foresight News16m ago

Robots Begin to 'Consume Data': The Hidden Production Chain from Indian Data Factories to Billion-Dollar Humanoid Robots

Robots have started to 'consume data,' driving the formation of a new industrial supply chain focused on producing training data for embodied AI. Unlike large language models, which are trained on vast internet text corpora, embodied AI models face a 'data desert' in the physical world. This has created a massive demand for first-person perspective video data (Ego Data), captured by workers wearing cameras in places like Indian garment factories. Companies like Neocambrian AI are establishing 'data factories' where workers perform standardized tasks (e.g., sorting clothes, kitchen organization) to generate thousands of hours of video. Research, such as NVIDIA's EgoScale, demonstrates that scaling this human demonstration data predictably improves robot performance, particularly for dexterous manipulation. This has validated a training path combining large-scale human data for pre-training with smaller amounts of robot-specific data for fine-tuning. The value of different data types varies significantly, forming a 'data pyramid.' The base consists of low-cost, large-scale internet and Ego Data. Higher layers include more expensive motion-capture data (e.g., from data gloves), simulation/synthetic data, and the most costly and scarce layer: real robot teleoperation data. This demand has spawned a layered ecosystem of data suppliers: low-cost data factories, motion capture and alignment specialists, robot-native teleoperation service providers, simulation data companies, and platforms aiming for data standardization. Robot companies themselves are adopting a 'layered procurement' strategy: outsourcing generic Ego Data while building in-house capabilities for robot-specific adaptation data and the critical deployment/failure data generated in real-world applications. The industry is shifting focus from hardware and basic mobility to the data pipelines required for general-purpose capability. While parallels exist to data labeling companies like Scale AI in the LLM boom, the physical complexity of robot data—involving action success ambiguity and sim-to-real gaps—requires more integrated solutions for data collection, annotation, and a continuous feedback loop. The race is on to build the data engines that will teach robots to operate reliably in the unstructured real world.

marsbit2h ago

Robots Begin to 'Consume Data': The Hidden Production Chain from Indian Data Factories to Billion-Dollar Humanoid Robots

marsbit2h ago

Spicy Commentary | Michael Saylor's 'Player Talk'; 60-Year-Old Aunt Liquidated After 'Scamming a Young Man'

**"Spicy Commentary": Three Tales of Crypto's Wild Week** This week's "Spicy Commentary" column highlights three dramatic stories from the cryptocurrency world. First, **MicroStrategy's Michael Saylor** addressed the controversy over his company potentially selling Bitcoin. At the BTC Prague event, he clarified, "I never said the company can't sell Bitcoin. I told *you* never to sell *your* Bitcoin." This "do as I say, not as I do" stance was criticized by netizens as peak linguistic gymnastics, noting a history of him previously stating the company would "never" sell. Second, a **bizarre fraud case** emerged from Beijing. A 60-year-old woman, obsessed with getting rich from crypto but unwilling to risk her own savings, posed online as the 20-something "god-daughter" of a high-ranking official. She catfished a young man, convincing him to give her over 200,000 yuan for fabricated emergencies. She then invested all the stolen money into cryptocurrency with 10x leverage, only to lose everything in a market crash. The woman was sentenced to four years in prison for fraud. Finally, a **sobering trader's tale** surfaced on Reddit. A user posted "Tale of a crypto trader," confessing their net worth had plummeted from a peak of $45 million to roughly $17,200, primarily due to holding meme coins too long. The post, described as a crypto "book of confessions," sparked reactions ranging from sympathy to critique about greed, poor risk management, and the perils of treating meme coins as long-term investments instead of taking profits. The column concludes that this week featured masterful rhetoric, elaborate scams, and extreme financial volatility, stitching together another chapter in crypto's unpredictable theater.

Foresight News3h ago

Spicy Commentary | Michael Saylor's 'Player Talk'; 60-Year-Old Aunt Liquidated After 'Scamming a Young Man'

Foresight News3h ago

Tremble Humans, AI Continues Its Accelerated Sprint

Trembling, Humans: AI Continues Its Accelerated Sprint Yes, AI is still rapidly accelerating. While deep learning seemed to stall quickly in its early years, large models after years of development show no sign of hitting their ceiling. At the Zhiyuan Conference 2026, the focus is on enabling AI to move from the digital world into the physical world. Scaling Law remains effective, continuing to drive advancements in both large language models and multimodal models. The industry is now entering a phase of pursuing World Models, though unresolved technical paths and data issues mean this exploration may take 3-5 more years. Concurrently, breakthroughs in Agents are accelerating AI's real-world application in fields like healthcare and meetings. Making Agents truly useful requires key hardware-software co-design, evident from the strong presence of chip vendors at the conference. We stand at a new historical threshold where AI is becoming a foundational force reshaping the world. The first day of the conference highlighted AI's evolution from "knowing how to chat" to "knowing how to work." Scaling Law persists, World Models are the next key battleground, and Agents are transitioning from usable to好用 (user-friendly). Scaling Law is not ending but diversifying. New models like Anthropic's Fable 5 demonstrate scaling through parameter size, synthetic data, and reinforcement learning. Advancements in AI Coding and Agent deployment are enabling a trend of AI self-evolution, potentially allowing AI to take over digital world iterations. World Models represent the next frontier for large models extending into the physical realm, but no current model is truly impressive at solving real-world problems. Technical consensus is lacking, with debates on data sources (video, simulation, real-world). Different approaches are emerging: language-centric, pixel-centric, 3D-structure-centric, and visual-representation-centric models. Zhiyuan Institute is exploring a fifth path: unified latent space modeling fusing language and visual representations, and introduced its own under-development World Model, Physis-v0.1. On the product side, Agents are key to bringing AI into daily life. Since 2025, the "Year of the Agent," products have become more proactive and capable of complex tasks. Zhiyuan showcased four vertical Agents for cardiac diagnosis, autonomous research, meeting summarization, and protein risk discovery. However, technical challenges remain, particularly in context engineering like memory and orchestration. "Harness" – the engineering framework around an Agent – is crucial for maximizing its capabilities by clarifying intent, designing workflows, and incorporating validation and feedback. In summary, AI's breakneck pace continues on multiple fronts: foundational model scaling, the ambitious pursuit of World Models for physical understanding, and the ongoing refinement of practical Agents. The journey from capable to truly reliable and useful AI systems is well underway.

marsbit3h ago

Tremble Humans, AI Continues Its Accelerated Sprint

marsbit3h ago

Trading

Spot
Futures

Hot Articles

Discussions

Welcome to the HTX Community. Here, you can stay informed about the latest platform developments and gain access to professional market insights. Users' opinions on the price of AI (AI) are presented below.

活动图片