Jeff Olson, Head of Applied AI & Analytics for ANZ at Cognizant, explains how evolutionary AI will change the way we manage businesses, problem-solving and decision-making.
In the current reality, taking a step back and looking at the big picture can be daunting. The way we work has fundamentally changed and digital transformation efforts have been supercharged out of necessity. Between both of these factors is the exponential rise in customer expectations and their increasing reliance on digital experiences every day.
In the middle of this storm, business leaders are expected to make increasingly high-stake decisions with speed and precision.
But faced with so many complex challenges, how can we ensure we keep steering in the right direction?
Enter Evolutionary Artificial Intelligence (E-AI). Behind these three words lies a completely new approach to leveraging artificial intelligence (AI), which has the potential to transform the way we manage businesses, problem-solving, and decision-making.
No one could have predicted what impact the pandemic would have on businesses and society, but what if we had better models for prescribing solutions to such an event? Traditional deep learning approaches have been inherently limited in their predictive capabilities, but E-AI technologies are making it possible to discover entirely new solutions — all on their own.
At this point, you might be thinking, “Isn’t this just one more thing to learn in an overwhelming deluge of new technologies?” Not really and here’s why. E-AI, by being faster and more insightful in the understanding of information and how machines learn, is already helping businesses in sectors ranging from manufacturing, finance, healthcare and agriculture achieve far better outcomes than traditional approaches to AI. With this in mind, let’s dive into how this is all possible.
E-AI: The DNA of better decision-making
According to Gartner, ANZ organisations are leading in digital maturity but falling behind in enterprise “fitness,” with many utilising digital technologies, like regular AI, to optimise existing business models, instead of preparing for new challenges or potential crises. This approach leaves a business vulnerable and stagnant. E-AI strategies represent an opportunity to create a competitive edge by taking AI to another level.
The way E-AI works was inspired by evolution. When facing a problem, an E-AI system will start generating a “population” of potential solutions to fix it. Then, it will evaluate each of these solutions, and the best ones will produce offspring through mutation, replacing the bad “candidates”. This process will repeat until a population of solutions that can solve the problem emerges.
One of the key differences between predictive (“regular”) AI and Evolutionary AI is the capacity for E-AI to come up with solutions never created before, where predictive AI offers the best possible solutions based on what it learnt from past models and situations.
Another key advantage is E-AI’s malleability. As the variables around a specific business decision or issue keep changing, the E-AI system constantly adjusts and produces new solutions that factor these new elements in. With predictive AI, the engine would have to re-process the whole situation all over again every time a variable changes.
Think of a business as a car. The way we currently leverage AI is similar to providing insights (GPS route, weather, distance when parking, indicators, etc.) to the driver to enable decision-making. E-AI, however, is an automated driving system that analyses all the variables surrounding the car that are likely to impact the driver at different stages of the journey. Based on this information, it makes the best-suited decisions to bring the car to its destination in the safest and most efficient way.
In business terms, an E-AI engine maps out your organisation, its challenges, situations and all external variables impacting it now and in the future, and prescribes entirely new models for business functions aimed at overcoming challenges while enhancing performance. From all possible outcomes that E-AI will offer, decision-makers simply choose the one they deem best.
All of this may sound a bit like science fiction, but E-AI is a reality and concrete applications are already starting to emerge. In Japan, the front of Maglev bullet trains have been designed by applying this method, which allows the manufacturers to factor in all relevant variables such as air resistance, speed or temperature, to come up with the most efficient design.
E-AI is also making waves in the financial services sector. For example, loan underwriting processes typically rely on human analysis to assess risk—but this can be a slow process and prone to human error. With E-AI, underwriters can analyse all combinations of relevant variables and contextual information faster to generate a model that more accurately assesses the risk of a potential borrower defaulting on their loan. Accurate assessments of loans also help to mitigate potential bias in the underwriting process – something traditional AI models struggle with.
Furthermore, E-AI allows companies to iterate in a safer environment. Currently, when an organisation tries something new, it has a direct impact on its employees, customers and other stakeholders. With E-AI, organisations can test every decision on a virtual representation of their business—a digital twin—and see what the impact would be on different elements of the structure.
As we look to constantly optimise every aspect of the organisation, solutions like E-AI that provide an overview of all the possible outcomes and the ability to choose the best one, are fundamentally changing the way business leaders make decisions. And in a rapidly moving world, those who can tirelessly iterate will naturally create a competitive edge. Where predictive AI has proven how it can help companies achieve that outcome, E-AI proves how it can do so on behalf of companies. Evolution is the essence of our world, and E-AI not only represents the next big thing after deep learning, but also an entirely new world for AI.
Jeff Olson is the Head of Applied AI & Analytics for ANZ at Cognizant.