Morgan Stanley, a leading financial services company, is set to roll out an OpenAI-powered chatbot to assist its 16,000 financial advisors. The advanced tool, which has been in development for a year, will help the bank’s advisors to access the bank’s vast data and research resources. The chatbot is based on GPT-4, a more advanced version of the technology that powers OpenAI’s ChatGPT.
The Promise and Perils of AI
Morgan Stanley’s move is among the first announcements by a financial services company since the success of ChatGPT, which generated human-like responses to questions. The technology has sparked a race among tech giants to develop generative AI products. However, the use of such technology has raised concerns about its reliability and accuracy. Last month, Morgan Stanley analysts reported that ChatGPT sometimes generated wrong answers, even though they appeared convincing.
User Guardrails for Accurate Responses
Similar to ChatGPT, the new tool will provide instant answers to advisors’ questions. However, it will only generate responses based on the 100,000 pieces of research that Morgan Stanley has vetted for this use, reducing the risk of errors. The bank will also have humans check the accuracy of responses to further reduce mishaps.
Breaking the Platform
Jeff McMillan, head of analytics and data at the firm’s wealth management division, is confident that the new tool will work effectively. He said, “We’re trying to actually break the platform through human testing. With high-quality information, better models, and an ongoing monitoring process, the bank is confident in its new tool.”
No Replacement for Human Empathy
While McMillan believes that technology can disrupt industries for routine tasks, he acknowledges that machines cannot replace human empathy when it comes to catering to sophisticated clients. “These things don’t have any empathy; they’re just very clever math that can regurgitate knowledge,” he said.