Rise of the Machines, but are they the Terminator or R2D2?
AI is big news, from driverless cars to the virtual assistant on your phone and no wonder, McKinsey estimates up to $39 Billion was invested in AI in 2016 alone. As Aaron Oliver, Director, Next Money commented "We live in a world of crypto currencies, smart contracts, and networks of things which enable machines to supply services, trade, negotiate, and transact directly. AI adds the layer of intelligence required for a machine to understand crowd psychology, to “feel" market sentiment, and thus elevates market evaluations beyond the data and technicals for all connected things.”
Much has been made of what this might mean for jobs. Machines already do many tasks that until a few decades ago required manual labour and it seems likely that more industries will be disrupted. FinTech is already seeing its own disruption from AI so are the machines out to get us? Or are they here to help?
It’s worth mentioning that AI has a broad definition, I’ll be covering that in more detail over the next few weeks. For the purpose of this article I’ve looked at 5 areas in which AI is transforming financial services. I’ve considered what it might mean for jobs in those fields as well as what potential the technology has to disrupt the market for customers.
Virtual Banking Assistants
What’s happening? Virtual Assistants, sometimes known as Chatbots, are used by banks and other FS firms to allow customers to access a range of services (make a transfer, pay a bill, check your balance etc.) without having to speak to an employee (either on the phone or in branch), think Siri but for banking. They are usually designed to recognise a variety of inputs from text to spoken word and Natural Language Processing aims to make this interaction as fluid as possible and make the experience ‘more human’. The goal is to minimise bank staff contact time and thereby reduce team sizes and cost whilst also creating a better customer experience.
Job risk 4/5
Disruptive potential 2/5
Verdict: Banks have long had large customer service and branch teams. With challenger banks able to remove this cost from their balance sheets incumbents are being forced to catch up. These teams are likely to be heavily reduced over the next few years. A genuine job Terminator. For customers though, beyond the novelty value, the difference will be fairly minimal with the same services likely to be available.
What’s happening? Until the advent of Robo Advisors, financial planning was typically the preserve of the wealthy. Most human advisors require clients to have a minimum of $100,000 of investible assets and charge between 1% and 2% a year. In 2008 Betterment launched the first Robo Advisor, these services use algorithms to invest assets for the client and involve no human interaction, saving on labour costs. They democratised the field with some services having no minimum account level and fees typically half to a quarter of human advisors.
Job risk 2/5
Disruptive potential 5/5
Verdict: The good news for advisors is that the target market for Robo Advisors doesn’t necessarily overlap with their core market so job losses should be minimal in the short-term. For customers though, high quality investment advice is far more accessible at a fraction of the price.
What’s happening? Traditional credit scoring takes in a range of factors to determine the level of risk in lending to an individual or business. AI led scoring uses machine learning to pick up on connections that previous algorithms couldn’t. Social credit scoring, of the kind China recently announced, take this a step further looking at a broader set of information (often from social media sites) to improve accuracy even more.
Job risk 1/5
Disruptive potential 3/5
Verdict: Credit scoring has been computer led for nearly two decades and statisticians are still required, to make sense of AI scoring and define parameters. For customers, especially those with no credit history, new scoring methodologies are likely to result in more accurate risk pricing. With the Chinese example who you are friends with on social media may soon affect everything from the price of a loan to how likely you are to be upgraded on a flight.
High Frequency Trading
What’s happening? Using algorithms in trading is nothing new and up to 80% of trades are done this way. However, algorithms are not necessarily AI although strides are being made in this area and most large funds now include AI as part of their trading strategy.
Job risk 1/5
Disruptive potential 2/5
Verdict: If your job security hasn’t already been threatened by an algorithm or AI in this space then chances are you’re safe. A mix of computer and human input is likely to continue for some time and whilst advances may be disruptive within their own industry for most of us they will have a minor impact.
Risk Analysis and Monitoring
What’s happening? Just as AI driven credit scoring is allowing lenders to be more accurate, AI is also helping Insurance firms to price risk. AI allows huge data sets to be used in analysing a variety of different risks, additionally, smart sensors, telematics and integrated systems allow insurers to monitor that risk in real time and mitigate some risk with customers, resulting in lower claims and better premiums for customers.
Job risk 2/5
Disruptive potential 4/5
Verdict: Underwriters will increasingly find themselves working alongside AI to understand and price risk for clients and since computer based models have been part of this process for decades, large scale job losses are unlikely. For customers, the ability to review behaviours and understand how they’re affecting premiums may well improve not only premiums but health and safety.