Ask HN: How to Fix Tech Hiring? | Hacker News Reader

Ask HN: How to Fix Tech Hiring?

4 pointsCharlieDigital9 comments

A partner of a friend recently took a few months off and then dedicated 30-45 days preparing for interviews with several Big Tech companies.

This individual basically dedicated this time to researching common leetcode problems (and cranking them several hours a day), common system design questions, and common behavioral questions for each of the target companies. Then some time was spent doing "practice" interviews by opening up their LinkedIn and effectively doing random throwaway interviews to prepare for the Big Tech interviews.

In the end, this process paid off and they got offers from a handful of companies and accepted an L6/staff role.

Now I am undergoing the same process in an ultra-condensed timeline and really lamenting the process. In a chat with the individual to prep for my own process, it was shared that most of the information had already been forgotten (not even started the role yet). Indeed, as I've been working through leetcode problems myself, I find that even some problems I solved just a few days old already require some effort to recall the solution. They also mentioned that the system design questions were the hardest and the one that they focused the most on because they didn't have much experience with system design (yes, read that again).

But perhaps somehow more troubling is that even as I crank and solve these problems, I do not feel that they have much real-world applicability at all and do not make me a better, more capable, or more productive engineer; it seems but a useless game to be played.

At a startup, early on we let go of one of our least productive, least capable, and least autonomous ex-Amazon engineers only to learn later that they ended up at Netflix a few months later...

The industry settled on such an awful paradigm. Thoughts on how Big Tech can fix this? Are there better paradigms? Better approaches or tools to evaluate engineers? I think startups and small teams have a bit more flexibility (the same friend relayed that another friend's startup interviews ONLY using an AI coding assistant as a pre-requisite), but what about Big Tech?

Goodhart's Law: "When a measure becomes a target, it ceases to be a good measure."

If one takes a step back, it's quite ridiculous and apparent the system is really irrational. Why would you ask a new grad about system design? How many new grads have actually had to deal with millions of users?

fwiw from my experience, the really strong new grads have either done relevant projects back at college or can come up with reasonably good answers on the spot (my team does a lot of pure algorithm/modeling/distributed system work so those are actually practical questions).

I do agree leetcode makes those questions significantly less useful

This is where the signaling comes in. Those grads almost certainly didn't have a real need to solve it, but knew that this was a signal companies were looking for, so they trained for it. So there was some correlation with effort and grit. The is a reasonably strong argument for leetcode/system design etc that what you're looking for in grads is grit/effort/energy, because they have very little experience.

I wonder though, if this is also why we see these cyclical rounds of massive layoffs. From personal experience, I know multiple individuals who took this route and most were good but not exceptional engineers in real life. None of them recent grads. All of them not particularly excited at the prospect of actually working at Big Tech, but getting paid and almost all mentioned "rest and vest" at some point in the conversation..........

I don't think Big Tech can be fixed, due to its Bigness. Size inevitably generates bureaucracy and bureaucratic behavior. Big companies want to do mass hiring and mass firing. Employees are considered interchangeable cogs in the machine. I'd be happy to learn of a counterexample, but I've never heard of one.

The only hope is that smaller companies resist emulating the big companies. The myth is that BigCo hiring practices are what spurred BigCo success, but the reality is that this myth has the cause and effect reversed.

I thought it was really brilliant when this friend shared that another friend hiring for a startup required the use of an AI coding assistant. Makes total sense in a startup context where code quality is second to speed and iteration.

But as these tools get better, where even Pichai is boasting of Google's use of AI in their codebase, I wonder how this will affect interviews at Big Tech and how they select candidates.

Big tech doesn't want to fix it. It has been completely taken over by MBAs/HR/make line go up types, who love their little PowerPoint slides showing how outsourcing a team to India will save 60% of costs. To them engineering quality is a very intangible metric, it always loses out to tangible metrics like cost and headcount. The only real way things change is for startups to overtake big tech by being more aggressive, smarter and faster.

A new generation of startups poaching employees from big tech needs to arise for tech hiring to be "fixed" (in favor of developers)

Reply on news.ycombinator.com