Every banking and financial institution will be looking into this for themselves. They’re sitting on huge LLM datasets that can be used by a GPT for better risk management, personalised financial products, customer service and more.
Why would you not take advantage of customer data, spending behaviour, debit, and credit patterns and then give this to the customer via a banking app for them to use individually as a personal financial guide (note: guidance, not advice!)
This could be a prime example of using artificial intelligence for the right reasons — giving power to people for their benefit. Check an example of this released this week — BloombergGPT: A Large Language Model for Finance.
A 50 billion parameter language model that is trained on a wide range of financial data. Construct a 363 billion token dataset based on Bloomberg’s extensive data sources, perhaps the largest domain-specific dataset yet, augmented with 345 billion tokens from general-purpose datasets.
It’s not open-sourced for obvious reasons!
Here’s the full paper to understand how they’ve built it and what they intend to do with it.