Episode Summary: “Artificial intelligence (AI) can be seen as a progression in our scalability of labor.” This quote comes from this week’s guest, Naveen Rao, who received his PhD in Neuroscience from Brown before becoming CEO at Nervana Systems, which works on full stack solutions to help companies solve machine learning (ML) problems at scale. In this week’s episode, Rao speaks about certain domains in industry where he feels optimistic about machine learning (ML) making a difference in the next five to 10 years, providing interesting perspectives that include advances in the areas of agriculture and oil & gas.


Get it on SC

iTunes Badge



Expertise: Neuroscience and machine learning

Recognition in Brief: Naveen Rao, PhD, co-founded Nervana Systems with the goal of changing the way the world thinks about computation based on deep learning technologies. In August 2016, Nervana Systems was acquired by Intel for a reported $408 million. The purchase was driven by Intel’s plans to integrate Nervana’s machine learning technology into the silicon of Intel’s own chip design. Nervana continues to operate with its proprietary team out of the company’s San Diego headquarters.

Prior to Nervana, Rao was a researcher neural computation and artificial systems at Qualcomm, Assistant VP at ITG specializing in transactional cost analysis, and also served a number of engineer roles, including as Principal Engineer at W&W Communications. He received his PhD in Neuroscience from Brown University and a BSEE in Electrical Engineering and Computer Science from Duke University.

Current Affiliations: CEO at Nervana Systems


Interview Highlights:

The following is a condensed version of the full audio interview, which is available in the above links on TechEmergence’s SoundCloud and iTunes stations.

(1:41) In the next 5 to 10 years, what areas do you think will grow the most (in applying ML)?

Naveen Rao: “There’s a framework you can use to make these kinds of analyses. AI as we know it today…is the next level of scalability of labor, when you think about it…we have data scientists who make algorithms that do things people used to do, and make it scalable and accessible…

…In the next 5 or 10 years, I think healthcare is a big one…agriculture’s a great one, it’s really just about using what we have today much more efficiently, using land more efficiently, growing crops with higher yield on that land, using energy that comes from the sun more efficiently…these are problems that kind of have hard limits on them, we can’t farm more area, so basically anything where we have limited resources and we have to use them more effectively.

…on the enterprise side – which tends to leads to consumers in general – you’re going to find the same concept of personalization but at the company level, companies building personalized data solutions fro their problems. I think that can be across many different types of industries; you see Walmart, Target, all these companies have data scientists, trying to compete with companies like Amazon; ultimately they’re all selling consumer goods, but they’re doing it in a way that’s more targeted, at the right time, and at the right price.”

(5:54) What is it about agriculture that makes it ripe for that kind of disruption and why do you feel it’s going to grow any more than any other industry?

NR: “…If you look at agriculture, it’s trillions (of dollars) per year, these are basic needs of humanity…from a technical side, what makes it good for disrupting it with robots is you have a very specific task, and a very specific constrained environment, and you can perform that effectively by optimizing a robot to do so…something where you know exactly what the task is, you know why the inputs are going to be, you know what the outputs are going to be, you can make something that’s very optimized and you don’t need to use human labor to do it.”

(7:35) What other sort of robotic applications might artificial intelligence (AI) weasel into in the agriculture world?

NR: “…crop selection, matching crops to various environmental variables, basically getting better utilization out of our resource of land, today it’s done mostly ad hoc…can we make better types of crops, crosses that work in different environments, can we shuffle around where crops are grown into different environments and actually get better yields – these are questions that are very answerable now, and using machine learning and data science, we can really optimize that.”

(10:50) In terms of personalized experiences…with respect to enterprise leading this domain, did you mean AI, that applications in a digital environment is normally led by bigger B2B companies?

NR: “…There are clear ROI statements to be made by enterprises today, this is why we’ve chosen to go with a more enterprise-centric view of the world…companies are much more rational about it i.e. if this is going to help us on our bottom line, then we want it.”

(14:39) When you’ve been speaking to the investor community…what did you find as the sentiment for AI in today’s day and age and what were the common misconceptions?

NR: “…When we started pitching people two years ago, one or two of the investors knew what deep learning was, or they’d heard the word, they didn’t have a good sense for what it was; I think there was a lot of excitement then, because that was when DeepMind was bought by Google…fast forward a year and a half, we closed around back in May of 2015, things were very different then, you talk to an associate or partner and they know exactly what deep learning is, they know who the players are…I think today there is a lot of excitement and I think everybody knows it’s going to be huge, it’s almost a certainty at this point because the capabilities that deep learning has brought to many industries are going to revolutionize them.”


Big Ideas:

1 – The healthcare and agriculture markets, two of the largest sectors in the U.S. economy, are poised to leverage AI technologies – in particular deep learning – over the next decade and undergo transformation as an industry, including in operations and delivery models.

2 – Personalized technology for users is a primary challenge and driver in today’s AI business and consumer markets.