Musings on ways to use generative AI in finance have continued: A tech leader at Goldman Sachs believes that large language models and chatbots could enable better knowledge sharing than traditional software-management systems, which have often failed to live up to expectations.
A top Goldman Sachs executive suggests that an AI-powered chatbot could help the bank record, store, retain, and access institutional knowledge.
Chief information officer Marco Argenti sent an email to the bank’s engineering staff, seen by Fortune, where he laid out a vision for a tool that could transform how the bank stores and accesses institutional info.
“Within a corporation, most knowledge is not codified. It’s tribal,” Argenti wrote. “It resides in the minds of ‘experts,’ connected by an internal social network that takes years to master.”
That obviously has its downsides: It’s difficult to know where to find key info, and it can be lost completely when employees leave. However, a “ChatGS” system that allows users to add and query information via natural language could make it easier for it to be recorded, stored, and made accessible.
By removing the manual processes or specificity needed to update and search more traditional corporate knowledge management systems, this strategy could break down information silos. In that way, large language models (LLMs) are “a breakthrough in knowledge more than they are in productivity,” Argenti said.
Internal knowledge sharing is just one way that financial institutions are thinking about using large language models and generative AI:
For example, Bloomberg is working on a tool that would let its terminal clients quickly retrieve proprietary information and ask financial-analysis questions, Morgan Stanley is giving its advisors access to a tool that synthesizes its research, Swedish investment firm EQT programmed a chatbot to help its dealmakers benefit from its “Motherbrain” data platform, and fintech Klarna has deployed it externally for shopping recommendations.