Technology is changing at the speed of light. I am fascinated by the implications of tools like Machine Learning and Deep Learning in the area of Finance. Business decisions have become ever so complex due to multiple factors. Information is being developed ever so rapidly and decisions are made in silos. In order to fully harness the benefits of Machine Learning in Finance a few hurdles have to be overcome.
Culture Change – The introduction of advanced analytical tools is a change management initiative. People are creatures of habit. For example, on morning commuters typically got to the same locations, on the bus and/or train. Everyone has his\her spot. People get fixed into a pattern and must be pushed into something different. At times, setbacks or challenges can function as a deterrent toward a pattern or routine. The setback can serve as a stimulus to improve an individual/group. For example, autonomous vehicles are currently being introduced into society. Yet setbacks scare many people away from this inevitable change. Recently, an autonomous Uber car crashed into and killed a pedestrian. In that particular case the automatic break feature was disabled. An Instance like that scares people against letting machines fully control critical parts of their lives. However, people must realize that the number of human produced accidents is far greater. The number of human decision errors far outpace that of a machine making routine judgements. The true culprit is control. People don’t want to give up the control of driving a vehicle or making business decisions.
Democratization – The next critical piece to this challenge of automated business decisions is the democratization of software and data. The right tools must be easily accessible. One of the reasons Excel is such a ubiquitous part of every business of its accessibility. Microsoft Excel came pre-loaded on every Windows computer. It was and continues to be the dominant application in every financial sector. Automated business tools need to follow a similar path. In a sense automated business tools should be offered freely, so that it can be embedded into everyday business life. The second critical piece is data. A person can have all the right tools but if they don’t have the data the tool is useless. Organizations are typically segmented, fractured and siloed. The organizational structure pose challenges for centralized data. In order to leverage autonomous business decision tools, data centralization must occur.
Training – We’ve heard time and time again that there is a need for tech savvy workers in the areas of data science. I would take that a step further and say that every single job will require a greater technical depth than in the past. The training to use the new tools and understand new concepts will have to start in the industry. The financial industry will need to take the lead in implementing the processes and training for the future. Higher Education will need to follow through, but that ship takes longer to reorient. Curriculum changes can take years before it is approved. The financial industry can react more quickly by invoking the resources and training needed.
The assistance of machines in our business decision will help improve decision making. However technological assistance will never replace the depth of the human intellect.