Machine learning and artificial intelligence (AI) could have a significant impact on the finance industry, but potential consequences still need analysed before being implemented, according to the “Artificial Intelligence and Machine Learning in Financial Services” roundtable hosted by Bravura Solutions.
There was broad consensus at the roundtable that machine learning could be used effectively to take on laborious work that required precision and accuracy, freeing humans to do more interesting tasks where they add creative value.
Among the talking points of the discussion panel:
- In the short-term machine learning is likely to be used to improve operations and quip humans with better information to drive good client experience.
- There is a growing body of evidence to suggest that algorithms are better at exercising judgement than humans. To implement machine learning, businesses need to identify where machines are better to work and where humans add value.
- Representatives from larger firms agreed they would first look at how machine learning could improve operations. Operations needed to cope with new business flows and poor operations could cancel out the benefits of implementation.
- Smaller firms want to use AI to scale-up as start-ups could find machine learning help them disrupt and achieve scale cost effectively.
- Data sharing offered a personalised service that would be cost effective to more consumers in an area where machine learning could make an impact.
There are still drawbacks as Simon Clare, Bravura Solutions, said: “Imagine you built a robo-adviser giving people advice based on historical model training and the Financial Conduct Authority (FCA) comes along and asks, why did you put this person into these particular products? And you say. Well the training algorithm said to. It is not an answer the regulator really wants to hear.”
