The mandate applies across most departments, requiring managers to verify that a recruit can identify AI-generated errors before granting access to automated assistants. Engineers remain exempt, as their output undergoes rigorous peer review. Wang, a former Goldman Sachs analyst, noted that the policy has already curbed costs, with projected annual token spending dropping from as much as $20 million to roughly $4 million.
Beyond the financial impact, the shift has forced junior staff to engage more directly with tenured colleagues to solve problems. This change addresses a growing workplace phenomenon where employees rely on unchecked AI outputs, a practice critics have dubbed the use of "meat proxies." While some external observers argue the policy stifles innovation, Wang maintains that the primary goal is ensuring staff possess the foundational knowledge to oversee technology rather than simply serving as conduits for it.

Comments (0)
No comments yet. Be the first!