AI/Automation
3 Trends With Big Impact On Enterprise AI Adoption
Artificial intelligence is now an intrinsic part of our daily lives. We think nothing of seeing personalized product recommendations on Amazon or optimized real-time directions on Google Maps. The day isn’t far when we will be able to summon driverless cars to take us home, where Alexa would have already ordered dinner after checking inventory with our smart oven and fridge. That being said, enterprise adoption of AI has been more measured but is evolving rapidly to accomplish tasks ranging from planning, forecasting, and predictive maintenance to customer service chat bots, and the like. With more and more enterprise tasks being performed by AI, the question I repeatedly get asked by business leaders is “what next?”
I believe, there are 3 big trends that would help propel AI adoption even further within the enterprise. First, is the move towards simplifying the science of AI. At present, AI requires very sophisticated human resources, such as data scientists to build machine learning models, and computational linguistics professionals to write knowledge extraction applications. This restricts AI applications and innovations to a select few, and consequently limits the speed of adoption within the enterprise. But this scenario will not last long. Technology companies are building tools to automate tasks performed by these skilled individuals, thus enabling even a data analyst or business user to build AI applications. For example, Infosys Nia, a next-generation AI platform built for enterprise, consolidates several AI technologies – machine learning, deep learning, knowledge extraction, natural language generation, among others – so that an enterprise can use the right tool for each of its problems. And, because most functions are automated on the platform, it brings down the time, cost, and effort, of adoption and innovation within the enterprise.
The second trend is an effort towards auditability and ‘explainability’ of AI, especially machine learning. This is an extremely important area for enterprises that operate in heavily regulated environments or where decisions made by machines can be life-altering. Increased use of techniques like deep learning has made the task of exposing the decision-making process of these AI systems even more difficult. While Explainable AI (XAI) work is still in its infancy, enterprise AI platforms like Infosys Nia have started including auditability and basic visualization tools to take steps towards a system that doesn’t behave like a black box. It would be hard to imagine wider adoption of AI within the enterprise without it.
And the third trend is the quest for producing high quality analytics and predictions with a small data set. Most of the machine learning algorithms used in the consumer space rely on huge amounts of data to achieve desired accuracy. Getting a large data set for companies like Google and Facebook is trivial but most enterprises do not have this luxury for many of their problems. For AI to be applicable for a wider set of problems within an enterprise (or across enterprises of different sizes and maturity), techniques need to be created that allow for accuracy with limited data. In addition to new machine learning techniques, domain expertise can also be used to supplement pure data-based learning.
In a way, these trends are about changing AI to accommodate enterprise needs. But what are enterprises doing to accelerate the adoption of AI?
In my view, enterprises are still at an early stage of the AI journey, a stage that is fraught with challenges. These challenges stem from the human impact of AI, rather than the technology itself. Every revolutionary technology in history has drawn its share of naysayers and change resisters. At the dawn of the Internet revolution, a number of people thought it was a temporary fad. Similarly, many didn’t take the mobile wave seriously till it was too late. Likewise, in today’s era of AI, there are many who doubt its potential and others who fear its impact. Organizations must allay these concerns swiftly because ultimately no business can succeed without the support of its people. Most enterprises will find in AI an opportunity to amplify the capabilities of their staff; when AI takes over routine, repetitive jobs, the people who used to do them can focus on uniquely human pursuits such as innovation, creative thinking and problem finding.
Organizations that are sensitive to the human factor, go further faster. Because, the best way to overcome any resistance is by involving employees in the transition from the start, and sharing with them both the larger vision at the enterprise level and their own renewed/ reskilled roles within it. While it may not be easy to make the transition to AI, if done right, AI can help enterprises differentiate themselves. At full potential, AI can achieve impressive results for the enterprise – serve customers better, improve a variety of business and efficiency metrics, scale without adding headcount, and above all, provide the deepest insights into the organization’s data.
