Most businesses believe that Artificial Intelligence (AI) is poised to acquire cult status, and everyone is taking steps to prepare for it. Many have started doing the preparation groundwork, and some have even done pilots or have gonelive. No financial organization will say that AI will not disrupt both – the market and the business strategies.
What is making waves now are Machine Learning (ML) and Deep Learning (DL). ML is where computers are fed a lot of data and humans train them to do the right analysis, such as analyzing loan applications or shortlisting candidates for recruitment. DL is where computers are fed a lot of data and they determine the right answers by themselves, without human guidance. This is seen in voice analysis, gaming, etc.
DL is a subset of ML, and ML is a subset of AI. While AI is still on its way, ML and DL have already arrived. Gartner estimates that by 2019, application functions based on AI/ ML/DL will be pervasive in 90% of the enterprises globally, means that they would have become widespread within 2-3 years. Already a large number of bigtech companies like Microsoft, Adobe, Salesforce, Google, SAP, etc, have embedded AI technologies into their solutions.
One important difference with traditional technologies is that ML/DL require a lot of training and fine tuning rather than software development. So, the question to consider is who should drive the adoption – IT teams or business teams?The earliest use of these technologies is in customer facing applications including marketing, campaign management, call center service and support, cybersecurity, etc. Language oriented AI encompassing natural language processing, translation, speech to text, text to speech, sentiment analysis, etc, are the earliest applications of AI in the financial sector. Early adopters of chatbots are already facing the challenges of scaling up and of having deeper understanding of customer text.
Another early use of AI is visual processing. An early use case is insurance companies analyzing vehicle accident images to assess repair costs. Another use case is face recognition for authentication.Other applications of AI/ML/DL include enhancing productivity of employees, identifying intangible risks, generating code, testing code, optimizing code, etc.
Adopting these technologies is hardly a challenge. Rather the challenge to the financial sector is how to remain relevant against fintechs and bigtechs which are poised to dominate front-end financial services using AI. The biggest challenge, however, will be figuring out how to enable collaboration between AI and human intelligence+wisdom, keeping in mind that AI evolution will be exponential and human evolution will be linear. Any thoughts?
