Artificial Intelligence: The Elusive Prize

How AI will impact the future of business - and how to get in front of it – Part 1

by Alex Blount, Mark Danaher, Jonathan Ebsworth, Tom Lurtz December 2017

Introduction

Few topics garner more headlines today than artificial intelligence, and for good reason. Despite the fact tech luminaries can’t seem to agree if it is the world’s greatest existential threat or saviour, there is little debate about its ability to fundamentally transform everyday life and the business models that support it.

As global markets are now flooded with start-ups and large consulting firms peddling AI, it has never been more important for corporations to have thoughtful answers to a pair of basic questions which we will explore in this series of e-books: how is AI likely to impact my business and what should I be doing to get in front of it?

Data: The Foundation of AI

Data: The Foundation of AI

AI is a set of techniques to use information to enable machines to exhibit intelligent action. The starting point for any company’s journey into artificial intelligence, therefore, is data. Companies make numerious decisions each day with incomplete information as they haven’t put in place the mechanisms to leverage all of the data that is out there. And they don’t all have the knowledge and imagination to connect the dots between the data they have and the business decisions that can be enhanced.

The starting point for any company’s journey into artificial intelligence is data.

As we observe the state of AI today, we see the industry attempting to address these foundational data challenges by competing for the mantle of powering AI for the masses. Firms like Google, Facebook or Amazon have begun to provide valuable infrastructure that can be used by entrepreneurs and businesses to load their data and begin to build intelligent algorithms without having an in-house team of deep learning and cognitive science experts.

Barriers to Adoption

Barriers to Adoption: Challenges that need to be overcome while considering AI

While we are seeing more and more firms choosing to pursue AI-centric solutions, often spurred to action by changes in their competitive landscape or visionary executives, AI remains early stage from a corporate-wide adoption perspective.

To better gauge actual AI penetration, Infosys recently put together an AI Maturity Index by interviewing over 1,600 IT and business executives from global corporations.

Despite a majority of executives seeing AI as a longterm imperative, only 10% believe their organizations are fully maximizing the available benefits and capabilities of the current opportunity.

So what, exactly, is holding the remaining firms back?

Barriers to AI Adoption:

Employees Fear Being Displaced

People are worried about losing their jobs to computer. While this concern is entirely reasonable, and should be a topic of continued dialogue, the reality is that, at companies where we’ve seen AI implemented well, machine learning algorithms have successsfully replaced mundane, data-driven tasks while freeing employees up to do more value-added activities.

Barriers to AI Adoption:

We Have A Hard Time Trusting Machines

Adding to the concerns about AI is a deep-rooted fear of losing control. Ironically, one reason machine learning is so powerful – and that neural networks are so vital – is that humans are not very good at stating what they implicitly know. This makes it very difficult to code clear, comprehensive instructions for doing repeatable tasks, therefore introducing the need for algorithms that can learn through observation. Intriguingly, the flip side is also true. Machine-learning algorithms are much better at making complex decisions than in telling us why they made them.

Barriers to AI Adoption:

Algorithms Need to Be Trained With Lots Of Data

From a technical perspective, another important barrier to AI adoption is the interaction that’s required with algorithms before they are able to deliver value. After neural networks are built and algorithms are written or connected via API, they need to be trained. Just like a child born into the world, for algorithms to work effectively, they need time to learn. The difference is that while humans learn via their voyage through life by observing millions of interactions that are logged in our subconsicious, algorithms must be explicitly fed data in a controlled fashion.

Barriers to AI Adoption:

AI Is Not Flexible

One of the most powerful aspects of the human mind is its flexibility. In an instant, humans can shift the context they are operating in, seamlessly moving between driving a car, choosing a restaurant and answering kids’ questions. This general awareness of our surroundings – and the ability to solve a diverse set of problems based on a complex set of environmental conditions – is one of our greatest gifts. Unfortunately, this is also one of AI’s greatest blind spots.

In part 2 of this e-books series, we will explore how companies can overcome these challenges and build scalable AI competency within their organization.

Alex Blount

Alex Blount

Partner

Alex is a veteran Partner at the firm, joining Lodestone as Director in 2009. He leads key technology and advisory services including strategy transformation for clients across Switzerland. He has 20+ years of industry expertise in manufacturing and has spent much of his career advising top global organizations on their growth and operational strategies – with a focus on how innovative technology can enable competitive advantage for them.

Mark Danaher

Mark Danaher

Partner

Mark is a partner at Infosys Consulting and the leader of the firm’s disruptive technologies practice – which combines some of the brightest minds around digital, big data, artificial intelligence and automation. In his 25 years of consulting experience, Mark has advised and delivered strategic solutions to clients globally, with a focus on the retail, manufacturing, transport and logistics sectors.

Jonathan Ebsworth

Jonathan Ebsworth

Partner

Jonathan is an automation and artificial intelligence partner at Infosys Consulting. He has spent over 30 years developing strategies and programs to help large clients transform their operations. He is an experienced program manager, enterprise architect and software engineer. Jonathan is also one of the firm’s leading design thinking practitioners.

Tom Lurtz

Tom Lurtz

Associate Partner

Tom is a member of Infosys Consulting’s disruptive technologies practice in Europe and also leads the organization in Germany that focuses on digital transformation and AI. His mission is to help transform companies into digitally-centric organizations, with a focus on customer interactions, new business models and product portfolio optimization.

Pin It on Pinterest

Share This