But megacorps hire in such volumes that it's simply not possible to ensure any quality at all by any measure. It's all just random people that even perform worse when put in a corporate environment. So overall it's below average quality.

So they should continue to hire by a metric that doesn’t work? Not try anything different at all?

What different metric do you think would work better? Github? I have seen people get job offers by taking someone else's project, editing commit history to make themselves the author, and then push it to GitHub and talk about that project in their interview. Sure, sometimes on drilling deep enough you would know people are faking it, but a lot of time people would pass through.

I just mean to say that all hiring criterion have tradeoffs, and none is strictly better than the others.

Nope, don’t give me the “oh there is no better way so let’s continue using our shitty way” excuse. Citing another poor way of hiring does not make the existing one better.

And I don’t know what a new process should be. Google has the data from literally millions of applications and has decided to just ignore them or not to try different ways of hiring? Give me a break.

Or maybe "the data from literally millions of applications" has made Google realize that their current way is the best among all the others possible? Who knows.

Or one day when Google isn't doing well there is a sudden phase transition to say the current way is the worst among all the possible.

We assume somebody competent is looking at this, but for all we know they are using the same dumb myopic evaluation as they use for promotions. I mean the same people came up with both.

How have Microsoft, Apple, and Amazon managed to diversify?

At least Microsoft and Amazon have the same interview style as google. The questions might not be as hard, but the interview style is same. All sorts of DS and algos. Infact, one of my friends once went to a microsoft interview, where even the interviewer had a leetcode like website open on his browser and asking questions from there directly.

Everybody follows the reputational leader. You can say Google ruined tech interviews.

Google didn't invent whiteboard interviews. Microsoft was doing it way back in the early nineties.

>So they should continue to hire by a metric that doesn’t work? Not try anything different at all?

If they're making money, then it works, and they should stick to it.

That's their actual metric, not developer competence in itself.

Blackberry’s highest revenue quarters were around 2010. Three years after the iPhone was introduced. Profits are a trailing indicator. But even if you don’t buy that, Google’s net income was down last quarter. Revenue was up but traffic acquisition costs were up even more.

What if Apple had stuck to just selling iPods in 2007? Just like Blackberry, more iPods were sold in 2007 than ever before - the same year the iPhone was introduced.

What if Netflix had kept shipping DVDs or more recently didn’t invest in its own content years before other providers started pulling theirs?

>What if Apple had stuck to just selling iPods in 2007? Just like Blackberry, more iPods were sold in 2008 than ever before - the same year the iPhone was introduced.

Sure, but how are those things relevant? Those are strategic decisions, not decisions made by developers -- and don't concern what we're discussing, if the devs they hire are good for the company.

If somebody at Google C-level decided "let's close down Search and Ads, and just make kitchen juicers" their profits would also tank.

But that's not related to if the devs they hire are good for them. Unless their profit is hampered by dev output (and not one of the other reasons it can go down), their devs are fine for them.

Who do you think gets promoted within Google what are their incentives? If engineers slowly move up the ranks based on the same culture and incentives - creating products that are soon abandoned. The entire culture of Google is not conducive to the diversification they need as ad sales become less profitable.

The culture is shaped from the leadership. There's no requirement a company needs to promote developers to strategic thinking positions.

There isn’t a requirement but from the outside it looks like that’s what’s happening.

Making money is one metric. How about diversified revenue stream? Do you think they should be concerned about that?

They probably should. But it's not the concern of their developers.