R. Lee Coulter, one of the organizers of the 2016 World BPO/ITO Forum, is senior vice president of St. Louis-based Ascension Health, the nation’s largest nonprofit health system, and CEO of its shared services subsidiary, the Ascension Ministry Service Center. His focus on shared services, business process outsourcing, or BPO, and technology spans a 30-year career in leadership positions at companies, including General Electric, AON and Kraft Foods. Here, he talks to SearchCIO senior executive editor Linda Tucci about the current state of artificial intelligence (AI) applications in outsourcing and what’s on the horizon. This interview has been condensed and edited.
How is AI being applied to BPO/ITO today?
Lee Coulter: AI is not being applied to outsourcing today, in so far as we’re talking about the shared services and outsourcing industry.
It’s important to note that AI doesn’t actually do anything, [whereas] the world of shared services and outsourcing is the delivery of services. Shared service organizations and BPO organizations exist fundamentally only to process what falls out of the straight-through processing technology.
So, if you have SAP, or you have PeopleSoft, Oracle or JD Edwards, or whatever, those systems were built with the intent for a significant portion of services to be provided without human interference.
It’s only when there’s an exception of some kind, meaning that the person needing the service requires something slightly different, or the conditions within the fulfillment of that service are slightly different than the base system was set to handle, that it falls out of the machine and a human takes it and tries to figure out what the person was really asking for. It is here where AI will, over time, begin to have some impact on shared services and outsourcing.
So, AI applications, at present, don’t act. What is AI?
AI is a set of rules and approaches to digesting a corpus. A corpus is the structured and unstructured data that an AI digests for the purposes of establishing its conclusions and correlations, so it can be used as a prediction engine. The whole purpose of AI is really to have digested enough circumstantial history that it’s in a position to answer a question with some level of confidence for a hypothetical or a future need.
In the world of automation as it exists today, whether we’re talking about machine learning platforms, or self-healing platforms or autonomic platforms — an easy-to-use term is task automation, or intelligent process automation. These technologies are immediately accessible in the world of BPO because we’re able to build rule sets into them. And these platforms are able to access any digital source of information necessary to get the data they would need in order to make a decision about what to do next.
What’s next for AI applications?
One of the areas of focus now is, do we actually give AI the ability to act in the physical world, or is that too dangerous? Do you establish some intentional barrier? You can imagine the dangers of a system that is capable of rewriting its own rules, and you give it access to act either in a physical world or in the digital world. Movies have been made about it — most notably, Terminator. But, again, applicability to the industry today is low.
What about processes that require interaction with humans, such as search engines or the future of automated telemarketing?
A lot of focus in this space is on natural language processing (NLP). There are very public and very successful NLP engines out there — Google’s DeepMind, Apple’s Siri [and] Microsoft’s Cortana. But these are monstrous investments, right? Tens of billions of dollars are being spent in developing these NLP engines.
R. Lee Coultersenior vice president at Ascension Health
Very quietly, Google — it’s probably close to a year ago now — actually implemented its own NLP AI into the standard search engine. Within 48 hours, 20% of all searches were being fulfilled by the new engine. And it fulfills virtually all search requests today. So that’s another important component here of seeing AI in the world of outsourcing.
Now, this is not cheap stuff. And most of ITO and BPO became ITO/BPO based on cost. Sure, people want some innovation, and they want better service levels and that sort of thing, but by and large, the first requirement of outsourcing is cost. That’s the ticket to ride. Those are table stakes. So, the application of very sophisticated AI will not occur until there’s a business case for it.
Consultants on intelligent process automation — or its future state anyway — are saying it will displace 140 million jobs. Soon, they say, we’re going to see automation taking over knowledge worker jobs, and the whole concept of labor arbitrage is going to change: If you can implement the software on premises, why would you go to India for this work, with all the uncertainties and complexities involved in that?
Coulter: This work is already coming back. People are bringing services back. There’s a term that I coined called no-shoring, where people who had work onshore are sending it into a data center, people who had work offshore are sending it into a data center and it lives in the cloud. It’s been no-shored. Once you automate it, it lives only in cyberspace, which doesn’t have a shore. Now, most recent numbers that I saw said that over a 10-year time horizon, 179 million jobs worldwide will be moderately to severely impacted by automation. Based on my last three or four years, I totally believe that.
You don’t think that’s overstating the case.
Coulter: I don’t — and this is me personally. We looked at this space four years ago, and back then, there was more hype around it. I said, ‘You know what, guys? Let’s take another look at it.’ Eighteen months later, we took a new look at it and we were like, ‘Holy crap! Orders of magnitude step-changing capability!’ I’ve been given the privilege to see the technology roadmaps of the leading providers [in process automation] here, and I can tell you that the next 18 months are going to be way more significant than the last 18 months.