We had the opportunity to sit down with Sudhir Jha, Mastercard Senior Vice President and Head of Brighterion, to talk about AI and to learn his perspective on major AI breakthroughs in the years ahead.

AI has come a long way in the last couple of decades. Would you say in the last 5 to 10 years, it’s really moved forward quickly?

AI is not new. It has been around for more than 60 years and  has gone through multiple hype and bust cycles. It even went through a dark phase in the 1990s where most people said, “Okay, it’s overpromised, it’s not going to work.”

In the last few years, it has found legs in terms of research and industry adoption. Every company that is moving to digital is generating tons and tons of data that you then need to analyze. The problem is that you have this vast amount of data. How do you find the nuggets from that data in the most efficient way and not really analyze things that can be happening today, but also predict something in the future? That’s where AI is really brilliant.

Where are we now in terms of the evolution of artificial intelligence?

I think AI has done a really good job of doing massive data processing and finding those nuggets, whether it is commonality or anomaly or other areas. Where it has still a long way to go is with natural language processing. How do you understand the context and logic when people are speaking or when they’re writing an email? You can get some context from there, but there is still a lot remaining to be done.

Which industries do you see AI being adopted in first or more quickly?

There have been some industries already using AI for a while now. Of course, the tech industry, especially Google, Facebook and Amazon, have been pioneers in terms of adoption of AI for multiple business processes within that. Financial services were also early in terms of adopting AI. A lot of that was in the trading, but retail had to follow quickly based on Amazon being there as well. Insurance, they are doing the right things, too.

I think the next wave is going to come from healthcare and manufacturing – industries that actually have massive, massive opportunities. Even education for example, they have been somewhat slow to adopt AI because their problems are much more complex. It also requires a lot more domain knowledge to really understand where AI can actually make a difference.

What are the benefits for those industries?

One, they have not done this before. Second, they have really complex processes that AI can make really big differences in, and they have massive investment.

Manufacturing has billions of dollars spent in doing things on the floor, supply chain management, all those kinds of things that really can benefit from AI. I definitely think there is going to be a shift where a lot of those industries that have not been adopting AI will adopt in the next few years.

What do you think the big AI developments will be in the years ahead?

An overall improvement that will occur over time is more advanced natural language processing and linguistics; being able to understand context versus just words.

On the technology side, everything stands to get better, like predictions, recommendations and models. I think one area that’s also going to have a lot more development (and some companies are already working on it) is how do you work with smaller sets of data. Of course, if you have a lot of data to train, machines can do a good job, but not every use case is going to have that much data.

I also think AI is going to be more accessible to more people in the years to come. Today nobody thinks about the Internet as an unfamiliar technology with an unclear use case, but AI hasn’t reached that point yet. People don’t understand what it is, they don’t know how to use it. In the next few years, AI tools are going to continue to get better, to the point that working with AI will be just as easy and familiar as working with Excel. All the technical details of AI will be hidden, enabling people to just say, “This is my data, this is what I want to know,” and they’ll be able to get results. That’s a big improvement that will come in the near future.

What are some AI adoption challenges you’re witnessing?

In terms of adoption, things are looking pretty good. Three or four years ago, when I started talking to enterprises about AI adoption, I had to begin by explaining what AI is! When I talk to leaders today, they all understand—at least at a high level—what AI is and what it can do for their business. They also understand that it’s something they must adopt in order to stay relevant and competitive.

Today, the core challenges are really more about how do you adopt AI? How do you select use cases? AI can fundamentally change lots of business functions in any enterprise, whether it’s supply chain, demand planning or Salesforce optimization. So how do you prioritize? How do you actually know which part of your business is going to have the maximum impact using AI?

AI depends heavily on the data and processes you already have, which means the same problem one company is trying to solve will have very different results using the same kind of AI technology [for another company]. The struggle most companies have is they try some use cases based on a recommendation from a consultant or AI company. Then, they don’t succeed because their data wasn’t there, or their processes were already optimized. They give up, assuming AI isn’t for them.

What do you say to people who fear AI will replace them? Do you think AI will really take over humans’ jobs in the years ahead?

Like any technology, AI is going to disrupt the human workforce. Look, when we moved from horse-drawn carriages to automobiles, some jobs were replaced. When RFID (Radio-Frequency Identification) technology was implemented in toll stations, toll collectors were replaced. Wherever AI can be used to automate things or do things more efficiently, there’s going to be some shift in the job market. But when automobiles replaced horse-drawn carriages, the human workforce wasn’t eliminated. Instead, humans started producing and/or driving automobiles. The same thing will happen with AI. It’s creating new jobs such as writing AI models and devising different use cases for AI. So yes, some humans will lose their jobs thanks to AI. But there will be a lot more opportunities for them to be re-skilled and redeployed in different areas that AI will open up for them.

Visit our AI Innovators video series to see Sudhir Jha discussing other future developments of AI.