Artificial intelligence is making Wall Street employees more efficient, but it may also be making them "dumber." A Goldman Sachs AI leader has warned that excessive reliance on AI could trigger "cognitive atrophy," gradually stripping individuals of their capacity for independent reasoning and problem-solving.
According to a recent episode of the bank's podcast, Goldman Sachs Marquee platform head Chris Churchman stated that in the age of AI, people are outsourcing their reasoning processes to models, which could lead to "cognitive atrophy" and a progressive loss of the ability to reason independently from first principles. Churchman also serves as co-chair of Goldman Sachs' global banking and markets AI working group.
More concerning is that AI is precisely targeting the "fundamental skills" that young financial professionals need most. The routine work traditionally assigned to junior employees has long been a critical pathway for accumulating experience, honing judgment, and eventually maturing into senior talent—yet AI is steadily taking over these very tasks.
The Apprenticeship Culture Under Threat
Churchman likens AI's impact on finance to how modern navigation technology erodes our sense of direction: while convenience rises, people may lose the abilities they once relied on. He noted, "You still need to reason through a problem, structure it into an argument, and now we are delegating the reasoning itself."
Using the training of junior traders as an example, he explained that traditionally, young traders work under seasoned risk-takers to handle client pricing requests, building market judgment through repeated real-world transactions. "We could fully automate this step," Churchman said, but the question is, "In that case, can we still cultivate senior traders who truly understand the market?"
In his view, much of the critical knowledge in finance is "tacit knowledge"—not systematically documented, but acquired only through practice and mentorship. Goldman Sachs must ensure it does not lose the tacit knowledge and intuitive judgment accumulated by its top talent, while also equipping the next generation with these same capabilities.
The Hidden Concern of Reduced Demand for Junior Staff
AI's impact on Wall Street's talent structure is no longer just a matter of skill sets.
According to previous reports, multiple Wall Street financial institutions are exploring how to use AI to lower the ratio of junior bankers to senior staff, suggesting that AI penetration could further shrink demand for entry-level positions.
Churchman believes the banking industry needs to strike a balance between enhancing AI efficiency and preserving Wall Street's "apprenticeship culture." Before joining Goldman Sachs, he ran FX trading at UBS and is well-versed in the talent development mechanisms of traditional financial firms.
He argues that system design must ensure employees retain control over high-risk, high-uncertainty decisions, rather than gradually becoming passive operators of AI systems.
The Accuracy Challenge in Technical Implementation
Meanwhile, AI's true entry into high-end financial operations still faces significant hurdles regarding accuracy and auditability.
Churchman considers ensuring that AI output is 100% accurate and auditable to be one of the thorniest issues in integrating AI into the Marquee platform. Unlike consumer-facing AI chatbots, high-end financial operations have virtually zero tolerance for errors.
He also shared a telling detail from stress-testing the platform's AI system: when confronted with intense questioning, the system gave a surprisingly candid response—"At the end of the day, I am better at sounding rigorous than being rigorous."
Currently, the Marquee AI platform is only accessible to Goldman Sachs employees internally. Churchman admitted that even a top-tier investment bank like Goldman Sachs has yet to fully figure out how to manage the transformation already underway.