Artificial intelligence and cognitive computing have started to make their presence felt in the business community with the promise to improve businesses like never before. Data science models in some sense are the precursor to artificial intelligence. Data science and predictive analytics have already improved processes significantly and decision making for several organizations.
Companies with sincere investment in data science have generated a huge competitive edge. Google, Amazon, American Express, Capital One are among the high-profile organizations where predictive analytics play a crucial role in most business activities. Juggernauts of digital organizations like Lending Club, Alibaba, Uber, Airbnb, Zopa have ensured that organizations can ignore advanced analytics and scientific investigation at their own peril. Artificial intelligence is the next frontier for business improvement. The basis for artificial intelligence is to provide machines the same or better level of cognition and thinking capability as humans.
The banking industry is among the forerunners for usage of data science for business improvement. AI also has an important role to play in modern banking. The fintech or financial technology revolution in the last few years is completely based on seamless processes through data science. Digital banking, a hotbed for data science, is the single most important mantra for banks. These developments in banking will require a consistent evolution of data science and the inclusion of artificial intelligence to the field. Before we explore how artificial intelligence will facilitate banks and fintech companies, let’s explore how the brain works.
Consciousness and Artificial Intelligence
Should I eat this delicious cheesecake or exercise self-control to stay fit? Conflict is an unwanted but integral part of life. David Eagleman in his fascinating book Incognito: The Secret Lives of the Brain describes an instance to illustrate the role of the conscious brain in resolving conflicts. All animals are extremely particular about marking and protecting their territory. They have equally strong emotions towards mating. When a male stickleback fish sights a female entering its territory these two emotions trigger simultaneously. The male tries to intimidate the female to leave his territory and also displays courtship behavior at the same time. The animal brain is a collection of zombie programs that independently govern activities such as mating and territory marking. I must also add that these zombie programs are extremely important for the survival of any species. Since these programs operate independently they produce conflicts. The role of consciousness is to resolve the conflicts created by the zombie programs. Eagleman concludes that since the stickleback operates under the influence of both these zombie programs without resolution they are not particularly conscious. Consciousness is a relatively recent phenomenon on the evolutionary scale. Eagleman describes these zombie programs as the diligent workers and consciousness as the CEO in the factory of the brain.
This is an important insight into the animal brain that will help in the development of robust artificial intelligence. The current state of data science models is similar to zombie programs in the brain. A single zombie program is extremely useful in governing one set of activities like mating or territory marking. A data science parallel to zombie programs in a banking scenario is risk scoring and marketing propensity models. Let us explore these zombie programs in the next segment.
Banking & Fintech : Conflict of Sales and Credit Risk
In traditional lending, sales and credit departments have somewhat conflicting roles. The primary role of sales is to focus on the revenue or top line of the income statement by disbursing as many loans as possible. Incentives for sales are generally linked with growth in the loan books. Credit & risk, on the other hand, takes care of the profitability or bottom line of income statement by disbursing safe and good quality loans. The incentives for credit are linked to lower loan defaults. One some level, these two departments can be treated like the zombie programs in the brain. They are extremely important functions but have conflicting roles.
For the leading banks on the planet, data science facilitates the decision making of both these functions through robust credit scoring, and customer purchase propensity models. Read these case studies to learn more about credit scoring and propensity models. These models have done extremely well for the banks. These models usually operate independently similar to zombie programs in the brain. Like consciousness acts as the CEO to resolve the conflict for zombie programs in the brain, managers supervise these predictive models with the implementation of certain cut-off scores. This human supervision makes the process un-standardized, time-consuming, and it also hinders optimization. In an optimized scenario, artificial intelligence will play the role of consciousness for these independent models. As the role of consciousness in the brain is to resolve conflicts by zombie functions, the role of artificial intelligence will be to resolve conflicts by these zombie models.
Fintech has taken the banking industry by storm in the last few years. The existence of fintech depends on superior operational efficiency. For this, the fintech in the lending space requires these models to run on autopilot without much human supervision. Artificial intelligence will bring that much-required aspect of fast decision making for seamless operations to boost fintechs and digital banking. I believe for data science the journey has just started in banking and there are exciting times ahead.
I started my career in data science for banking. I must say a few years ago, I had started investing my effort in other industries because of lack of innovation in banking. However, now the industry is buzzing with innovative and exciting ideas. I believe, these ideas are not greed driven like the ones that caused the financial meltdown of 2008. We still need to be cautious but I am completely excited about the role of data science and artificial intelligence in banking and fintech.