Cliff Justice | 13 February 2020
A few years ago, most CEOs would have considered artificial intelligence (AI) a shiny new technology their business could experiment with. Such is the pace of AI adoption that today, in many industries, it’s becoming part of the business mainstream. Our report Easing the pressure points: The state of intelligent automation reveals that more than 30 percent of enterprises invest more than US$50 million in intelligent automation technologies.1 AI is not a single thing or piece of software. It is a probabilistic technology that learns from human interaction, intelligence and data to provide a new source of intelligence to software and machines.
AI’s transformative power is best compared to that of electricity at the start of the 20th century. AI, for its part, will play a similar role in the fourth industrial revolution.
Today, organizations are exploring three broad categories where AI promises massive change.
The first is Insights - making sense of complex patterns and behaviors coming from many sources of structured and unstructured data like language, video or photos. This insight can pick up weak signals or patterns and help predict and foresee patterns of behavior on a large scale.
The second is Augmentation - boosting human knowledge and capabilities by leveraging data to extend the available collective knowledge to the customer base and workforce. Already early examples of this are emerging through virtual assistants and visual augmented reality.
The third category is Automation - the ability to use technologies like deep learning and natural language processing to perform tasks and execute processes that previously required human capabilities to read, see and understand the human environment and make decisions in the context of policies or expected norms.
All of these examples of AI are in the very early days of adoption and very few have achieved any type of scale in the enterprise. Nonetheless, already it is clear that AI is a transformative, learning technology with massive potential for change – something the business world is quickly grasping.
As the KPMG 2019 Enterprise AI Adoption Study2 found, the world’s leading AI adopters already commit significant resources to AI. The five most AI-advanced companies have an average of 375 full-time employees working on AI, including data scientists, engineers, analysts and others. They anticipate their headcount to increase to between 500 and 600 full-time employees in the next 3 years.
Given the growth potential of AI, what lessons can CEOs draw from successful AI adopters? Here are five insights that can help you to navigate the AI adoption journey.
AI has the potential to be integrated into every part of your future business operations. Indeed, most enterprise technologies and systems of engagement now build AI into their core products as standard and half of the companies interviewed for the AI adoption study expect to be using AI at scale within 3 years.
The key to implementing AI effectively will be to identify the specific business needs that your existing technology or processes can’t solve. That will involve engaging many areas of the organization to identify exactly where AI can add real value because AI success rests on having the right organizational capabilities to develop and support new technologies. One of the encouraging findings of the study was that half the companies interviewed said their CIO has a leading role for overall AI strategy while 40 percent said a senior Line of Business (LOB) leader plays a leading role. Among the companies interviewed, there are multiple technology and line of business stakeholders with responsibility for AI strategy and deployment.
Evaluating your organization and approach to see that you have the right structure, leadership oversight and talent in place is essential for fostering the spirit and structure of collaboration that will create a pragmatic platform for setting realistic goals and timelines. That requires executives to treat AI not as a standalone technology, but as a business driver throughout the organization and clarity on why you want to embrace AI, whether it be for quality improvement, customer experience, growth or productivity.
The speed at which companies adopt AI will be a key differentiator in the marketplace. Managing the rate of adoption can be important in order to win the acceptance and trust of employees, customers, regulators and other stakeholders.
In our experience, companies should follow an AI adoption model that addresses the needs of the customer, while considering the enterprise technology architecture, processes and access to the necessary data. Take our work helping Transamerica understand and better utilize the massive amounts of unstructured data it collects from customer contracts related to the London Inter-Bank Offered Rate (LIBOR), the benchmark average interest rate banks use when borrowing from each other.
Using natural language processing technology to understand unstructured conversation we then created a speech-to-text prototype to convert call recordings into plain text and accurately reveal in seconds the context of each conversation. Finally, we built an automated consumer graph database to help Transamerica fuel more innovative decision-making, higher levels of customer service and greater process accuracy compared to manual effort.
Pilot projects are necessary for any AI adoption, not least to overcome basic objections to the validity of the technology and prove the potential. However, pilot projects can become a trap. That is because pilots typically take place in a very controlled environment with people that are invested in that project. As a result, a scaled deployment will look and perform differently and required much more data, change management and oversight than a limited pilot.
For AI adoption to move beyond the pilot phase, you must consider how the technology will play out in a business as usual context. How will your customers, employees, regulators and other stakeholders feel about it? Will it make their experience better? How does this impact other business processes and enterprise risk? Is the effort and cost to change worth the reward? Do you have the necessary leadership to drive this change?
The AI leaders already have firm plans to move beyond the pilot stage: half of the companies in the study expect to be using AI/machine learning at scale within 3 years and 83 percent expect to be doing the same with robotic process automation.
Successfully integrating AI into an organization is not just a technology issue – it’s also about talent. On average, the five companies in the study considered to be the most mature in their AI capabilities already are spending US$75 million on AI talent today.
The rationale for this investment is simple. True innovators within your business will help you shape the future of the organization. But it’s not as simple as hiring the right talent. These innovators also have to be nurtured. The executive team has to be open-minded enough to consider their ideas, nurture them, and utilize this fresh thinking to embrace AI and try out new ways of working. This can be a challenge in many organizations because the traditional working environment doesn’t lend itself to attracting and retaining the type of talent needed to make AI a success. Innovators must be happy in the workplace and feel their work has a sense of purpose.
Failure to properly value and develop innovation talent means your brightest and best may swiftly move on to a more inclusive working environment. They may be tempted by a tech firm or start-up that may turn out to be a future competitor for your business.
Nobody that is alive today has witnessed a change like the one AI will bring. As electricity did more than a hundred years ago, there is likely no part of our daily lives that AI won’t affect in some way.
Those companies already advanced in AI adoption have grasped the scale of change that is coming. In the AI adoption study we found a ten-fold gap in resources devoted to AI between the companies with more mature AI capabilities and those at the early stage.
To prepare for a business landscape unlike anything any of us have experience before takes vision and planning. Asking yourself this question: if you were running a startup digital version of your own business, how would you disrupt the status quo? Understanding that new way of thinking will help guide your decisions in AI investment.
One thing you can be sure of is that in this newly emerging AI-enabled business landscape, if you are not identifying ways to transform your own business model, someone may already be planning to disrupt it.
1 Easing the pressure points: The state of intelligent automation (PDF 2.1 MB)
2 AI tranforming the enterprise (PDF 5.5 MB)
For more insight on data-related topics, please visit our data-driven technologies article series page.