The future of finance will not be won by the firms that copy each other fastest. When I first wrote this piece in 2023, language models were new and every bank announced a chatbot. Today the tools are everywhere, and the real question is what to do with them.
What language models do well
Models such as GPT from OpenAI read large amounts of unstructured text and turn it into something usable. Therefore, they are good at summarising research, answering standard questions and filling in forms. In retail banking, that shortens waiting times. In wealth management, it removes hours of administrative work, for example in account opening. I welcome all of that.
Where the future of finance stays human
However, analysis is not advice. A model can find a pattern in a portfolio. It cannot know that a client is about to sell a company, or that a family is quarrelling about succession. Those facts come out in conversation, and often only after years of trust. As a result, the more routine work the machines take over, the more the personal part of the job is worth.
Conformity is the real risk
My worry is a different one. If every bank buys the same model, trains it on the same data and asks it the same questions, every bank will give the same answer. That is conformity with better software. For clients, it means less choice, not more. So the firms that stand out in the future of finance will use the tools to free up time. Then they will spend that time on thinking for themselves.
Revolutionise or regret was the old title, and it was too loud. The quieter version is this: adopt the technology, and keep your own opinion. The GPT revolution makes the first part easy. The second part, however, is still up to us.
Source: https://vapa.ch/navigating-the-future-of-finance-beyond-conformity/ · vapa.ch


