This article is classic 20/20 hindsight that you always see when a business decision doesn't pan out. From the author:
> The self-created tragedy of Meta is that the company is loath to invent new products. Instead, Meta’s management more or less relies on “vibes” to govern its decisions, and those vibes are often either wrong or far too late. The most pertinent example of the former is the ill-fated metaverse, which was developed solely on the (unbelievable) whim that the pandemic would last forever — or at least far longer than it actually did — and people would become accustomed to replacing in-person interaction with virtual reality. It was precisely at this moment, roughly around mid-2020, that Meta (then Facebook) disintegrated from a social media company into a Ship of Theseus that still technically operated its core social platforms but fundamentally was distracted by a red herring. Vibes-based management.
This is bullshit. All businesses need to take calculated risks, which sometimes don't pan out. The reason Zuck went full bore into VR/AR is because he desperately understood (correctly, IMO) that not owning a major computing platform was a huge risk. Microsoft had Windows on the desktop, Apple and Google had iOS and Android on mobile. When iOS restricted advertiser tracking (requiring opt in, which nearly every same person thought was a good thing) it cost Facebook a ton of money. So Zuckerberg made a big bet that VR/AR would be the next big computing platform.
Except the bet flopped. Turned out very few people want a big pair of goggles strapped to their face all day. But Facebook wasn't alone in making a bad bet - Apple spent billions on Vision Pro, and that's basically dead.
The problem isn't that Facebook made a big bet that flopped, and that alone isn't "vibes management". The problem is that since Zuck has majority voting power, basically all decisions come down to him and nothing is tempered by the expectations/desires of others, so you get this "pendulum" management as I call it. Sometimes you have these visionary leaders and it works out, but a lot of times it crashes over time. Musk runs his companies largely the same way, and while obviously it has worked out well for his companies and shareholders, it's widely known now that employees at his companies see a primary task as "managing Musk". Tesla's self-driving capabilities are far behind Waymo despite having a lot more driving data, and a big issue is Musk's insistance on cameras only. I've seen comments here on HN that any self-respecting AI leader would never work for X AI because the marching orders are basically "grok has to say what Musk wants".
Having an "emperor CEO" means the company can move quickly, but that goes for both beneficial and detrimental directions.