Things have never trended in our direction. You have to will the future into existence: NVIDIA's Jensen Huang on AI and the future of tech
The era of physical AI is here..and more predictions by Jensen Huang
In this Siggraph keynote conversation, NVIDIA CEO Jensen Huang discusses the company's journey from computer graphics to AI, the impact of generative AI on various industries, and NVIDIA's vision for the future of computing.
Huang talks about the challenges and opportunities presented by accelerated computing, the importance of open-source AI, and the role of NVIDIA in shaping the next generation of technology.
Big ideas from Jensen Huang’s talk
NVIDIA's Evolution: From Graphics to AI
NVIDIA's journey began with computer graphics in 1993, but the company's focus on accelerated computing led to breakthroughs in AI. The development of CUDA and the first contact with AI through AlexNet in 2012 marked significant milestones. This evolution culminated in the creation of DGX-1 in 2016, the first computer built specifically for deep learning. The company's ability to adapt and innovate in response to technological advancements has been crucial to its success in the AI era.
Generative AI: A New Paradigm in Computing
Generative AI represents a fundamental shift in software development and problem-solving. Unlike traditional programming methods, generative AI can learn from inputs and example outputs to create programs autonomously. This approach enables the solving of previously unsolvable problems and has applications across various fields, including scientific computing, creative industries, and robotics. The potential of generative AI to transform industries and increase productivity is immense, but it also presents new challenges in terms of control and accuracy.
"We now have machines that are learning how to program the software to writing software that no humans can, solving problems that we could barely imagine before."
Open-Source AI and Democratization
NVIDIA's commitment to open-source AI, exemplified by the release of Llama 3.1, aims to democratize AI technology. By making large language models accessible to a wide range of industries and companies, NVIDIA is fostering innovation and enabling broader adoption of AI technologies.
This approach contrasts with closed-source competitors and reflects NVIDIA's strategy to maintain its leadership position by driving industry-wide advancements.
Accelerated Computing: The Key to Energy Efficiency
Accelerated computing is crucial for managing the increasing energy demands of AI and data centers. By designing specialized processors and systems, NVIDIA can achieve significant speedups and energy savings compared to general-purpose computing.
This approach is essential for sustaining the growth of AI and data-intensive applications while addressing concerns about energy consumption and environmental impact.
“Accelerated computing helps you save so much energy 20 times, 50 times, and doing the same processing. So the first thing that we have to do, you know, as a society is accelerate up every application we can."
The Future of AI Assistants and Digital Agents
Huang envisions a future where AI assistants augment every job within companies. These digital agents, powered by retrieval augmented generation and connected to digital human interfaces, could revolutionize customer service and various other industries.
By capturing institutional knowledge and providing more engaging interactions, these AI assistants have the potential to enhance productivity and transform how businesses operate.
Physical AI and Robotics
NVIDIA is pioneering the concept of "physical AI," which involves creating AI models that can interact with the physical world, particularly in robotics. This complex process requires three distinct computing platforms: one for creating the AI, another for simulating it, and a third for running the AI in physical robots.
NVIDIA's comprehensive approach, including tools like RoboCasa and Isaac SIM, aims to simplify and accelerate the development of advanced robotics applications.
"The era of physical AI is here. Physical AI, models that can understand and interact with the physical world will embody robots. Many will be humanoid robots."
Energy Efficiency and Data Center Innovation
Addressing concerns about the energy consumption of AI and data centers, Huang proposes innovative solutions. These include relocating data centers to areas with excess energy resources and focusing on inference rather than training as the primary goal of generative AI.
By optimizing energy use and leveraging AI to improve efficiency across various sectors, NVIDIA aims to mitigate the environmental impact of increased computational demands.
"AI doesn't care where it goes to school. Today's data centers are built near the power grid, where society is, of course, because that's where we need it. In the future, you're going to see data centers being built in different parts of the world where there's excess energy."
The Importance of Open Standards
NVIDIA strongly supports open standards, particularly Open USD (Universal Scene Description), as a means of ensuring long-term accessibility and collaboration in digital content creation. Comparing Open USD to HTML for virtual worlds, Huang emphasizes its potential to unite various design tools and enable seamless collaboration across platforms. This commitment to open standards reflects NVIDIA's broader strategy of fostering an inclusive and innovative tech ecosystem.
Continuous Innovation and Adaptation
NVIDIA's success is attributed to its ability to continuously innovate and adapt to new technological challenges. From graphics to AI and now to physical AI, the company has consistently pushed boundaries and opened up new markets. This approach requires constant learning, reinvention, and a willingness to tackle complex problems across various domains, from healthcare to robotics.
"Things have never trended in our direction. You have to will the future into existence."
The Role of AI in Future Work Environments
As AI continues to advance, Huang predicts significant changes in how work is performed across all industries. While acknowledging that jobs will change, he emphasizes the potential for AI to augment human capabilities rather than replace them entirely.
The future workplace is envisioned as a collaborative environment where humans work alongside AI assistants, leveraging their combined strengths to achieve greater productivity and innovation.