Or if they did internally, why the Board decided to ignore the most loyal and proven.
Googler #1 to #1000 were clearly not fools, and by the 2010s they very likely had enough credibility to be taken seriously if several dozen of them said this.
this is beyond the plane of rationality and the social forces take on a life of their own. it's not a math problem.
Yes I have actually, several times.
It's not easy, and of course it gets exponentially harder as the group grow, but these folks also have access to exponentially greater resources and exponentially greater motivation to do so. (Assuming their desire is to remain loyal shareholders of the company for the long term)
Hmm this sounds like the concern was badly communicated because this phrasing is both unnecessary and not quite correct either. (since it is possible to keep recruiting to a very high standard even during exponential growth via spending exponentially more resources)
The straightforward logic of:
linear increase in organization size = exponential increase in organization complexity = exponential frictional losses
would have been much clearer.
Since when? You're assuming the pool of employees available to recruit is infinite which is clearly not true.
It will breakdown if the company wants to run multiple Manhattan Projects simultaneously, but anything short of that is technically possible.
I claimed it's technically possible with 'exponentially more resources', not that any private organization could feasibly possess the resources or cachet to, in practice, hire a million top notch people.
Maintaining the literal same average quality with Googler #1001 to #10000 would probably have required spending 10x more per hire, compared to hiring #1 to #1000.
I have a slightly different answer than most of the other replies, which is that Google just got too big. I worked there for 10 years (left a year ago for a smaller company) and the amount of inefficiency and waste due to having too many layers of middle management with no real visibility into what was going on was just astounding. Over that decade they paid me millions of dollars in total, while canceling whatever I was working on almost exactly every 18 months like clockwork. Typically due to management issues or corporate politics rather than for any real technical or business reason.