Valon CEO bars AI use for most new employees to boost skill development
Valon’s chief executive Andrew Wang has introduced a rule that prevents the majority of new hires from using AI tools until they demonstrate the ability to spot errors.
Valon, a New York-based startup that builds AI-driven mortgage-servicing software, has mandated that nearly all new employees complete their duties without artificial-intelligence assistance until they can reliably identify mistakes. CEO and co-founder Andrew Wang launched the rule last month after discovering that recent hires were defaulting to the costliest AI models even for trivial assignments, a habit he feared would both raise expenses and hinder genuine skill acquisition.
The restriction applies across most departments, including senior positions, with the exception of engineers whose code is always peer-reviewed. Wang says senior staff have largely supported the move, as they previously spent time correcting AI-generated errors from newcomers. The policy is expected to cut the company’s annual token spend to roughly $4 million-$5 million, down from $15 million-$20 million, while encouraging deeper understanding among recruits. Critics on social media argue the approach is misguided, but Wang invites alternative suggestions.
Why it matters
The rule highlights growing concerns about over-reliance on AI in the workplace and its impact on costs and employee development.
In this story
