Naveen Rao Drives Artificial Intelligence into the Future

by Deb Miller Landau
iQ Managing Editor

Intel’s head of AI talks about driving racecars, growing up in rural Kentucky and how AI marks a natural evolution of the human race.

Under the blazing sun at a race track in Thermal, CA, a pioneer in artificial intelligence grips the wheel of his Ferrari 488 and floors the accelerator. The tinder dry vegetation outside becomes a blur and the smell of sun scorched asphalt feeds his adrenaline rush. Naveen Rao, the driver, is fired up.

“I love building stuff and diving into the technical details,” said Rao, a neuroscientist and engineer who leads Intel’s Artificial Intelligence Products Group (AIPG) at Intel. “Race car driving is about skill, but also about complex engineering and figuring out how to make the car perform better. There’s a very technical aspect to it that I like.”

Racing cars is meditative for Rao; it helps him think. And he’s got a lot to think about. The multimillion dollar deal that brought him to Intel put him in charge of building artificial intelligence (AI) technologies that transform how businesses operate and how people engage with the world.

Like racing Ferraris, he knows winning requires speed, precision, focus and guts.

“I think our brain is the best computer out there because it takes in massive amounts of information, processes it and does something useful with it,” he said. “Artificial Intelligence is attempting to solve that same problem.”

Research firm Tractica forecasts that the revenue generated from the direct and indirect application of AI software is estimated to grow from $643.7 million in 2016 to $36.8 billion by 2025. This represents a significant growth curve for the 9-year period with a compound annual growth rate (CAGR) of 56.8 percent.

AI and precision engineering is all around today’s society – facial recognition on Facebook, the way Google can find answers almost instantly, how a student can find a research paper from 1985 in seconds and cars that drive autonomously. Many of these were thought impossible a few years ago, but are ubiquitous today.

“I really do believe that AI is the natural evolution of humanity,” said Rao. “It’s the natural evolution of intelligence.”

Whether it’s cacti surviving life in the desert, animals adapting for survival or humans able to coordinate information and make better decisions – Rao observes that when complexity in the world increases, living things adapt, be it the cactus or the beetle or the human.

Rao believes in an increasingly complex and interconnect world, even technologies need to learn and adapt. That’s the promise of AI.

Early Days

Rao’s family moved to a small town in eastern Kentucky in the 1970s, when Rao was just three. They came from India, by way of the UK. His father was the local pediatrician. “I don’t think they had any idea what they were getting into,” Rao said.

Rao had a poster of a Lamborghini on his wall and built transmissions out of Legos. He and his friends rode bikes and played in the woods. While Rao and his brother fit in with other kids, his family looked different and they were culturally different from their neighbors. They would travel, back to India and other parts of the world, giving Rao exposure to global perspectives.

“Most of my friends had never been outside the area, had never been on a plane and had no intention of doing so,” he said.

His Indian culture inspired a work ethic that propelled Rao and his brother to be excellent students, and later, valedictorians of their high-school classes. Rao was voted most likely to succeed.

Rao’s father was an avid tinkerer and he encouraged his sons to examine how things were made, to question why they work. He’d drive his boys 40 miles away to the Radio Shack because he saw their early interest in computers.

That interest in engineering persisted. Rao completed his undergraduate degree at Duke, followed by a master’s in electrical engineering from Stanford. He worked for a decade as a design engineer before going to Brown to get a PhD in computational neuroscience – just so he could fully understand how the brain works.

He is known as an over-achiever and risk-taker and insatiable boundary-pusher. His daring nature led him to bicycle racing, mountain biking and racing in the Ferrari Challenge, a race series where each driver races the same model car, a Ferrari 488. Rao finished second in his debut at the 2017 race at Laguna Seca.

Naveen Rao races in the Ferrari Challenge.

“He has an insatiable curiosity coupled with strong intellect and personal discipline,” said Matt Ocko, co-founder of Data Collective, a venture capital fund that invests in seed and early-stage Big Data and IT infrastructure companies. “When Naveen wants to learn about something, he masters it relentlessly, in a way that enables him to both offer value and leadership to others.”

In 2014, along with pals Amir Khosrowshahi, and Arjun Bansal, Rao co-founded Nervana Systems, which created AI tools to help businesses develop custom deep-learning software. Intel acquired the startup in 2016 for $408 million. Rao went from having 50 employees to more than 800 today. And Intel continues to invest in AI technologies. Intel CEO Brian Krzanich said the company is investing a billion dollars in the AI ecosystem to fuel adoption and product innovation.

“We started Nervana with the goal in mind of actually changing the world of computing,” said Rao.

Ocko believes Rao’s capable of doing it.

“Naveen combines the confidence that comes from real intellectual firepower, years of deep study, and repeated success, with the humility of a truly committed scientist who delights in – and can respond to without fear or bias – the unexpected and disruptive,” said Ocko. “He is genuinely an ethical and thoughtful person, which engenders trust in both the teams he has led and the people who work with him.”

Moving Forward

Rao’s garage in Thermal is spotless, with two shiny Ferrari’s and a replica of a Shelby Cobra that Rao built in college. He walks past and notices a small puddle underneath the Shelby – a leak of some sort that will have to be repaired. He shrugs, knowing this is part of it – understanding how all the pieces fit together so when something leaks, he can figure out how to fix it.

In the same way engineering continues to evolve to make racecars faster and more efficient, technology is evolving to make AI a seamlessly integrated part of life.

“All of a sudden we have enough data, Moore’s Law has gotten to a point where we can put enough devices on a chip and we want experiences that learn from our behavior,” said Rao. “The world has just changed. We’re at the very beginnings.”

At Intel, Rao and his teams build the silicon chips and software tools for developers to build and scale AI. He wants Intel computation technologies to power the evolution of AI. To do this, he leverages Intel’s vast range of technologies, ranging from microprocessors designed for low power devices and chips that power big data centers, to customizable integrated FPGAs (field-programmable gate array), 5G wireless networking and IOT solutions.

He points to oxen pulling a till – that oxen-and-till application boosted productivity and changed farmers’ lives. Rao believes that building the tools and level of computation required for AI applications will continue to change lives in the future.

“If you look at humanity, I think we have a very important role to play in the evolution of our species,” said Rao. And like the farmer who no longer had to till by hand, jobs will shift and change.

“I think AI is going to create a lot of new opportunities for people,” said Rao, adding that some jobs will go away but new careers will open up. AI will take up mundane tasks in factories of the future making way for data scientists and analysts.

Top universities already offer programs focused on computer science, cognitive psychology and engineering. Rao knows the next generation will build the future for the next, and he’s already seeing that in action.

“I sat on a plane with a bunch of kids from Stanford, and they were all talking about containers in Hadoop and convolutional neural networks, and I was like, ‘What?!?’,” said Rao. “I mean this stuff was not even high-end research when I was in school.”

Back on the track, as Rao takes practice laps to get ready for his next race, he goes to his Zen place. The buzz of the engine, the screeching tires, the thrust and jolt of adrenaline as he hits a straightaway – it’s all a marvel of engineering, risk-taking and pushing boundaries. This is Rao’s happy place.

 

Editor’s Note: Click here to learn more about Intel technologies advancing Artificial Intelligence.

Share This Article

Read Full Story