Podcast: The tech underpinning the AI revolution

AI experts and their opinions seem to be growing on trees these days, but what’s harder to come by is insight into the computing tech that powers AI models. If there’s one person who knows something about that, it’s Jensen Huang, CEO and founder of Nvidia.

Huang was recently a guest on the No Priors podcast, hosted by investors Elad Gil and Sarah Guo.

At the risk of getting too meta, we used the AI chatbot ChatGPT to summarize this podcast about AI tech. The following is an edited version of its summary:

In this episode, Huang discusses the history of his company and its work in accelerated computing, including its role in artificial intelligence.

He explains how Nvidia initially focused on accelerated computing when the industry was mostly concentrated on general-purpose computing. He also gets into how the company’s work moved into self-driving cars, robotics, climate science, digital biology, and eventually, AI.

Huang details how the company expanded the flexibility of its accelerators and invented a new computing model called CUDA, which became an important tool in AI research. He also touches upon the importance of Nvidia’s GPUs in the current AI ecosystem and how they have enabled the development of new models and applications.

Huang, Gil, and Guo also discuss the evolution of AI technology over the past decade, including the development of reasoning, perception, and language models such as AI transformers and BERT.

They talk about how the language of programming has been completely disrupted because, for the first time in computing history, a computer can now be programmed using any human language instead of with a coding language like HTML or Python. This has a wide reach and is the reason why ChatGPT is the fastest-growing application in history, according to Huang.