‘A computer would deserve to be called intelligent if it could deceive a human being into believing that it was human.’ said Alan Turing.
The question of the moment is how Artificial Intelligence (AI) can improve retail financial services.
Companies have invested heavily, so it is unsurprising that executives want to push AI-driven products. ChatGPT has opened Pandora’s Box and suddenly everyone is a blue-sky thinker.
However, surveys show that consumers do not trust AI, partly because it lacks empathy. So, the sector has work to do if it is to be more than just an adjuvant technology.
To improve the entire caboodle: customer acquisition, product, sales, and support, the industry needs to explain how AI supports advisers but does not replace them.
There are three stages to achieving this: define today’s technology; be honest about consumer preference; specify steps to accelerate greater acceptance.
Firstly, the nomenclature is misleading: current AI is not true Artificial Intelligence. It does not pass the Turing Test, fooling a human into thinking that there is another human and not a machine at the end of a connection.
Instead, it refers to Machine Learning (ML) where computers use algorithms to parse vast data sets and make deductions.
In retail financial services it can be linked to pattern recognition, link analysis and optical character recognition (OCR) to perform KYC.
The advantages of this are greater pace, scale, accuracy and lower process costs, the disadvantages are that it is expensive to implement, opaque, the cost of making a mistake could be higher and it might infringe human rights by blocking certain customers for the wrong reasons.
Also, this technology does not engage consumers like humans. Therefore, changing the moniker from AI to ML and explaining how the technology is used would reduce consumer anxiety.
Secondly, humans enjoy talking to humans. Baby Boomers (1946-64) and Generation X (1965-1979)- 80% of UK wealth- prefer consulting an adviser. Both seek a proven track record, superior product advice, and a client relationship.
Later groupings -Gen Z and A – described as digital natives – are readier to embrace ML but they only ‘slightly trust’ it.
These latter groups have high expectations about new technology, and although always willing to try it out, will drop the brand if it disappoints.
They also value humans by relying on influencers and peers to make decisions. So, for all cohorts, companies need to reemphasise the value of their advisers.
Thirdly, retail finance requires the empathy of human connection. Part genetic, but mostly passed on from parent to child, scientists describe empathy ‘as a complex capability enabling individuals to understand and feel the emotional states of others, resulting in compassionate behaviour. It requires cognitive, emotional, behavioural, and moral capacities to understand and respond to the suffering of others.’ (Dr Riess, PubMed).
A lot of work has been done to recognise and label human moods, but technology still struggles to understand complex emotions. It cannot read the gaps between what people say they feel and what they actually feel.
ML capable of empathy is an exceptionally long way off, but it should be used to prompt contact with customers. Seamlessly switching from ML-driven analysis to human contact with customers would empower better service.
In conclusion, ML can improve customer processes and drive greater efficiency but only advisers are able to bridge the empathy gap.
The value of a retail financial services brand is still in its advisers and their ability to build customer relationships.
Matt Smith is chief content officer at WHJE and managing director of WPB