NEW YORK — A Goldman Sachs executive leading one of the investment bank’s major artificial intelligence initiatives is warning that Wall Street’s growing reliance on AI could weaken the analytical skills of the next generation of bankers and traders.
Chris Churchman, a Goldman partner who leads Marquee, the firm’s digital platform for institutional clients, told CNBC financial firms risk allowing employees to outsource too much of their reasoning to AI models.
“There’s a huge danger here that in the era of AI, we outsource our reasoning to these models, and we have cognitive atrophy that stops us being able to reason from first principles ourselves,” Churchman said on an episode of Goldman’s “Exchanges” podcast, according to a transcript provided to CNBC.
‘Reasoning is Still Important’
Churchman said AI could have an effect similar to technologies that have reduced people’s reliance on navigation and memorization skills. For Wall Street, however, the concern is that algorithms could increasingly perform the analytical work traditionally used to train young bankers and traders.
“Reasoning is still important,” Churchman told CNBC. “You still need to reason about [problems] and structure it into an argument, and now we’re delegating reasoning.”
Wall Street firms have been moving aggressively to integrate AI into trading, investment banking and other operations, potentially increasing productivity and reducing costs. But CNBC reported that the shift also raises questions about how firms will develop future senior employees if AI eliminates much of the routine work historically performed by junior staff.
‘A Need for Balance’
CNBC reported last year that Wall Street firms were examining whether AI could allow them to reduce the ratio of junior bankers to senior employees.
Churchman, who ran currency trading at UBS before joining Goldman in 2021, said banks will need to balance automation with preserving the industry’s apprenticeship model.
“You learn by doing, and a lot of knowledge is tacit, it was never written down,” he said.
For example, junior traders traditionally develop skills by responding to client pricing requests while working under experienced traders. While AI could automate those tasks, Churchman questioned whether doing so would eventually leave firms without senior traders who have developed a deep understanding of markets and risk.
“We can absolutely automate that,” he told CNBC, “but then do we get the senior traders that fully understand?”
Churchman told the news outlet financial firms need to design AI systems so employees continue making decisions when stakes are high and outcomes are uncertain rather than simply becoming operators of automated systems.
Goldman itself has not yet determined how best to manage that transition, according to Churchman, who also serves as co-chair of the firm’s Global Banking and Markets AI working group.
