Why Jensen Huang and Marc Benioff see ‘gigantic’ opportunity for agentic AI

Going forward, the opportunity for AI agents will be “gigantic,” according to Nvidia founder and CEO Jensen Huang. 

Already, progress is “spectacular and surprising,” with AI development moving faster and faster and the industry getting into the “flywheel zone” that technology needs to advance, Huang said in a fireside chat at Salesforce’s flagship event Dreamforce this week. 

“This is an extraordinary time,” Huang said while on stage with Marc Benioff, Salesforce chair, CEO and co-founder. “In no time in history has technology moved faster than Moore’s Law. We’re moving way faster than Moore’s Law, are arguably reasonably Moore’s Law squared.”

Agents working with other agents, ‘working with us’

In the future, Huang noted, there will be AI agents that understand subtleties and that can reason and collaborate. They’ll be able to find other agents to “work together, assemble together,” while also talking to humans and soliciting feedback to improve their dialogue and outputs. Some will be “excellent” at particular skills, while others will be more general purpose, he noted.

“We’ll have agents working with agents, agents working with us,” said Huang. “We’re going to supercharge the ever-loving daylights of our company. We’re going to come to work and a bunch of work we didn’t even realize needed to be done will be done.”

Adoption needs to be demystified, he and Benioff agreed, with Huang noting that “it’s going to be a lot more like onboarding employees.”

Benioff, for his part, underscored the importance of people being able to “actually understand” how they work and their purpose, and “need to get their hands in the soil.”

“Building an agent should not be some computer science fair project,” he said. 

Still, Huang pointed out that the challenges we have in front of us are “many.” Some of these include fine-tuning and guardrailing, but scientists are making advancements in these areas every day. In an interesting feedback loop, AI is being used to curate data to create a safe curriculum to teach AI. 

“It’s now reasoning about ‘Is the answer I’m generating sufficiently safe and proper, and is it the best possible answer I can be providing?” Huang explained. 

Nvidia ‘did a couple things right’

Early on, Huang explained, Nvidia observed that general-purpose computing would be good at some things but not others and that there would also be “interesting problems” to solve that would require some computing augmentation. 

The company then focused heavily on accelerated computing architecture, augmenting CPUs with GPUs and building out its DGX platform. “We knew that if we wanted to be a computing platform, we had to be architecturally compatible,” said Huang. “The fundamental tenant of the company was selecting problems that this computer architecture could solve.”

He noted that “all kinds of complex algorithms” were ported into Nvidia’s computing platform Cuda, and the company began to leverage deep learning. One of their early observations was that “deep learning would change software altogether,” said Huang. “We had the conviction to re-engineer every single stack of computing as a result.”

Nvidia had the advantage, Huang noted, of “working with every researcher on the planet.” They observed early on (in 2011) scientific work to train one of the first larger computer vision models. 

“The breakthrough was when we realized that unsupervised learning was going to be possible,” he said. 

Ultimately, humans would be limiters of digital AI because it’s impossible for us to label at scale, he pointed out. Instead, scientists are using language models to create other language models with multimodal data. That feedback loop is advancing at an “incredible rate.” 

“We knew today was going to come all along,” he quipped, joking that “we called it to the day.” In reality, though, he acknowledged that “we did a couple things right.”

Benioff agreed, saying that “in my wildest dreams I never thought [accelerated computing] could do what it can do now.” 

What motivates Huang and Nvidia?

When asked about his personal motivation, Huang described a tangible excitement. “It’s within your grasp,” he said. We can do this. We can make a real contribution.”

He added that he’s “sufficiently humble” and understands that he doesn’t know everything; lifelong learning is essential. 

“When you learn something it gets you fired up,” he said. “When you connect to random ideas that nobody realized could be connected, you get fired up.”

Nvidia and others will ultimately bring a level of automation capability that the world has never seen, he pointed out, saying his company is in a once-in-a-“lifetime position and a once-in-a-generation position.”

He marveled: “Right now it’s just too thrilling, don’t you think? Nobody should miss the next decade. You’re not going to want to miss this movie.”