Scale AI Founder and Meta Chief AI Officer Alexandr Wang: Hands-On Control Was Once One of the Key Success Factors, Then Became the Bottleneck
Alexandr Wang shared on a podcast how his management approach shifted from Scale AI to Meta. Early on, personally controlling every detail was one of the important factors behind the company's success, but when every decision had to pass through him alone, he himself became the bottleneck. After joining Meta, he was neither an AI researcher nor a product designer, and with the lab facing extreme time pressure, he pivoted to recruiting top talent and letting specialists shape the direction.
ImportanceMaterialEvidenceE2 unreplicatedWrite-upQuick
Wang's advice to first-time founders is: when starting your first company, the founder should personally take on every role, deeply understand every detail, and only gradually learn to delegate. That advice itself is not new, but he immediately followed it with a specific turning point—at Scale AI, personally controlling every detail was once one of the important factors behind the company's success, but it later turned him into the bottleneck for business progress.
The way this bottleneck formed was quite specific: every decision had to pass through him personally. When the company was small and he could judge every link himself, this centralized control brought speed; when the complexity of the business exceeded what one person could simultaneously manage, the same centralized control became a constraint.
After joining Meta, the situation he faced did not allow for a gradual transition. The Superintelligence Lab needed to complete a large volume of work in a very short time, and he himself was neither an AI researcher nor a product designer—he could not personally handle all the key work the lab required. The path of doing things himself first and then gradually delegating did not work here, because the conditions themselves did not permit it.