This wide-ranging interview with Jensen Huang, CEO of NVIDIA, covers a multitude of topics from the company's history to the future of artificial intelligence and its impact on society. Huang reflects on NVIDIA's early days, recalling how the company delivered the first supercomputer to OpenAI back in 2016. He discusses the paradigm shift in computer science brought about by deep learning, highlighting the need for entirely new computer architectures tailored to handle the massive data sets and complex algorithms required for AI.
He stresses the immense computational demands of AI, explaining that models like GPT-3, with its 175 billion parameters, require entire data centers to run effectively. This complexity necessitated a complete rethinking of computer design, from processors to networking. Huang discusses the acquisition of Mellanox, a pivotal move to enhance computer interconnectivity and enable millions of GPU cores to work in tandem. He then explains the concept of multimodality and how GTP-4 is able to learn from language and images, giving the example of horses and zebras to clarify that having the ability to compare one thing you know well with something that has been described allows you to use your knowledge in ways that you couldn't if you didn't know both.
Huang asserts that AI is reaching its "iPhone moment," democratizing computer programming and closing the technology divide. He believes that the barrier to programming has been reduced, allowing billions of people to participate and contribute. He predicts a significant boost to productivity across knowledge-based professions, as AI automates mundane tasks and empowers individuals with access to vast amounts of information. The interview shifts to the societal implications of AI, forecasting a new industrial revolution driven by the production of intelligence itself. He acknowledges that some jobs will be displaced while new roles will emerge.
He anticipates that climate change research and drug discovery will be two areas profoundly impacted by AI in the next five to ten years. He foresees AI facilitating the engineering of proteins with specific functions, as well as accelerating the drug discovery process. He addresses the ethical concerns surrounding AI, acknowledging the potential for misuse and emphasizing the need for regulation. Huang supports governmental oversight of AI, drawing parallels to the regulation of other potentially harmful technologies. He expresses concerns about the potential for AI-generated fake news and its destabilizing effects. He says that putting all of that power into the hands of one group is dangerous, so efforts should be made to democratize access to AI.
He delves into the characteristics that have contributed to NVIDIA's success. Huang attributes the company's resilience and agility to the numerous challenges it has faced throughout its history. He emphasizes the importance of a forward-looking vision, a willingness to learn, and the character of the people within the company. He speaks of the importance of the corporate culture and that that culture cannot be captured in words on a document but needs to be lived and breathed through the people who work for the company. He reiterates the importance of ingenuity, tenacity, and willpower in overcoming obstacles. He describes his upbringing in a family where it was hard to discourage them or distract them.
Finally, Huang shares his advice for young people, advocating for thinking from first principles and not relying on the advice of others. He encourages them to fall in love with what they do, rather than seeking out something to love, as this fosters a deeper commitment and passion for excellence. He describes how he's been in love with every single job he's ever had.
The interview concludes with Huang emphasizing the potential for AI to reshape the world and that he has personally changed the world to such a great extent.