Got big data? Embrace AI innovation and empower your people to answer better and harder questions, advises Jeff Wong of EY.
With almost a quarter of a million employees spanning over 700 offices in 150 countries, EY is an organization that literally spans the globe and touches a multitude of customers and clients throughout the professional services landscape. One might assume that such a large organization (listed by Forbes magazine as one of the top ten largest private companies in the US for 2017) would struggle to embrace innovation and explore new technologies such as artificial intelligence (AI) and automation. Think again.
To get a better perspective on what EY is doing with AI, we talked with Jeff Wong, the organization’s Global Chief Innovation Officer. Jeff recounted some of the successes and opportunities he’s experienced enmeshing AI innovation into the company’s culture, and he also shared what he has been hearing from clients on the challenges of creating an AI-enabled workplace.
“In the first 90 days we were 200% to 300% better than the current process across every relevant metric. We didn’t spend very much time, but we found a massive return. This woke me up to the power of artificial intelligence.” – Jeff Wong, EY
ANSWERS: What are some of the most common questions your clients are asking about artificial intelligence?
JEFF WONG: At EY, we have 250,000 people around the world and we serve clients across pretty much all industries, every sector and geography, big and small and everything in between. What’s been really nice about my role is that I’m able to hear from client groups around the world and of all different sizes.
What are the clients really saying about AI today? They fall into two groups. One is unsure of where to start. They know it’s important and they’ve read a bunch of things, but they’re not sure what to do. Then there’s a second group which is starting to experiment with this technology, but they are having a hard time finding people to work on it.
We conducted an artificial intelligence poll of 122 business leaders at the EmTech Digital Conference, produced by MIT Technology Review Insights this past April, and 28% of the respondents said their organizations basically have limited or no capabilities in AI. I put those respondents in the bucket of “unsure of where to start.” What’s interesting is the rest of the group, (more than two-thirds of the group) said they were starting to experiment with AI but they can’t extend their scale or find the talent. They want to make more investment, but they can’t find the talent.
ANSWERS: What is the biggest misconception you regularly encounter that clients are saying about what AI can do for a business and on the other side, what is the reality?
WONG: The biggest misconception out there is about the impact of AI. The framing out in the world seems to want to lean towards the negative: “AI is going to replace jobs.” There is a sense of apprehension and fear around what AI will do to people’s time and the availability of great jobs in the economy.
At EY, we don’t see it that way at all. In fact 99.9% of the organizations I interact with don’t see it that way, either. The way that EY looks at it is that we need to give our people superpowers, and AI technology and tools is a fantastic way to give them superpowers, to extend and scale-up what they’re capable of doing. In other words, to be able to ask and answer harder and better questions.
Our people are motivated to ask for more AI-enabled capabilities. If we didn’t do this for them, they’d be demanding it from us. In all of our investments around the world and in each of our service lines as well as at our global lab, people are clamoring to be involved with the technologies that we’re putting out there; one of the cornerstone technologies that we’re investing in is artificial intelligence.
ANSWERS: What are the hallmarks that a company is approaching innovations such as artificial intelligence effectively and maximizing its investments?
WONG: Measurement and understanding what you return is very important for an innovation group anywhere. Particularly, when it comes to technologies where you’re making new investments and you’re less certain about the outcome, it is important to measure the impact – to be honest about things that are working well and should receive more investment versus those that aren’t working well and should receive less or no investment.
This past year, we’re measured our savings in the couple of millions of hours. We measured the number of bots we deployed in the 2,000-ish range. For every single one of the bots that we deployed, we looked at how much we invested and how much we got back.
Advisory is one of our big service lines, and we’ve grown our robotic process automation (RPA) group and have a $750 million pipeline. It was a technology we started really focusing on about 18 months ago, and we weren’t sure how big it could get. But we’re pretty confident and have done pretty well. But when we started with it, it was a better-known technology than say AI or blockchain.
Regarding AI, one of the very first experiments that my team got involved with was training a bot to read contracts. As you can imagine, EY as an organization reads a lot of business documents. It’s a lot of the time spent in what we do – reading, analyzing, and understanding business documents. Thinking about them and trying to get to a conclusion or an answer. Of course, we read a lot of legal documents. The easiest one conceptually that we thought about for our AI project was the typical lease contract because it’s a fairly structured document. Somebody’s renting a building, and the data behind that are how much they rented it for and the terms behind it. We read hundreds of millions of pages of lease contracts every year. In fact, the number might even be in the billions but where we’ve been able to account for it, it’s in the hundreds of millions of pages at least.
That takes a lot of time, and so what we said is, “Maybe we can train an AI bot to read some of these documents.” To be honest, I thought, “Well, this is a little bit of a scientific experiment. It’s going to take us a little while to get this technology working. I’ll give it 12-24 months. That’s probably a reasonable time to see return.” However, in the first 90 days we were 200% to 300% better than the current process across every relevant metric. We didn’t spend very much time, but we found a massive return. This woke me up to the power of artificial intelligence. I didn’t think we would have performance at this level, and in fact I thought we were going to have a third of the performance we ended up having.
The fact is there’s impact from these technologies today so you can measure them and figure out where you need to get better and how you need to get better. We measure everything. The milestones might be different, the targets might be different, the hypotheses might be different but we make sure we measure everything because that’s the way we know how to figure out how to improve.
ANSWERS: What excites you the most about artificial intelligence development in the next three to five years?
WONG: The technology itself is just a fascinating technology. There’s a lot more progress to be made, and I think a lot of progress that will be made in the next three to five years. I’m very excited about this technology and its implication and potential, but there are two things I’m probably most excited about.
One is how early we are. People like to talk about artificial intelligence as one monolith thing, but it’s really a series of different technologies, some which are more developed and some less. There are people working on these problems around the world from large research and development and big technology companies to startups to academic institution labs. We’re going to make incredible progress over the next three or five years along different core components. It’s just a really fun thing for me to think about the possibility of what’s going to be available to us beyond the immediate financial impact that we have seen today.
The second thing I’m really excited about is that there are large pools of data that exist in the world that allow competitive advantage in terms of the problems you’re thinking about. In our industry, we process a lot of financial documents and transaction information around the world, in every major country and pretty much every major city. We touch every sector and a large number of companies small, medium and large, and the amount of data that crosses our desks every day is incredible.
To me, I am just really impressed by our ability to take that data, match it up with AI technology and the promise that it brings, and to develop outcomes where we can help companies make better decisions about how they act, operate, and decide what’s important and what’s not.
Explore more of our new series, AI Experts, where we interview thought leaders from a variety of disciplines — including technology executives, academics, robotics experts and policymakers — on what we might expect as the days race forward towards our AI tomorrow.