Comment by yodsanklai | Hacker News Reader

Comment by yodsanklai | Hacker News Reader

Parent Full thread

> An ex-FAANG worker in the right job might be able to be a 10-100x force multiplier, just by building the right tech to make other developer's job simpler

I don't think your random FAANG worker could show up somewhere and just "build the right tech" on their own, especially in a not so good team. There are some very strong SWEs, but the average one isn't a super star. My experience working at a FAANG is that everyone in my team is good (able to solve problems, hard working, good foundational CS knowledge), and there may be 1-2 very strong ones who really make things happen. Also, a lot of the skills we develop are with our own internal systems, I don't know how easily it would transfer to open source world.

In my previous company, there were some good engineers, but also some mediocre ones. This led to a lot of bikeshedding, wasted time trying to convince people, to review bad codes and so on. What I'm trying to say is that a very good engineer may not be able to unlock their potential in bad team, unless they also have good leadership skills.

The top 1% of SWEs in any FAANG are world-class engineers and are unique. I'm talking about the John Carmack's of the world who are top of their field.

The top 10% at FAANGs are game changers, and having them in any company would probably accelerate them. I've worked with some really extra ordinary, high-output engineers that were brilliant.

The rest of the 89% are very, very ordinary. I'm at a FAANG and the vast majority are exactly like every other SWE I've worked with in my career. There's nothing special about them. I'm currently working with some of the laziest, dumbest SWEs in my career despite their employment at a FAANG.

So to those that think FAANG is some badge of honor, don't believe the hype, I don't think FAANG employment is a meaningful signal at all.

> So to those that think FAANG is some badge of honor, don't believe the hype, I don't think FAANG employment is a meaningful signal at all.

I honestly think FAANG employees sank their own boat. The constant bragging of how little work the do in a week on anonymous forums, and the "day in the life" videos that show everything but work. Despite their bragging about being 10x, I don't see any proof of it.

The 1% of SWE though are very good (especially older employees with years of experience e.g. Rob Pike). The rest of the 89%, as you say, are mediocre at best. A Google employee in 2010 is not the same as a Google employee in 2020.

The 1% who brag are rarely SWEs IME - usually PMs and admins and the SWEs who do rarely do it for long generaly just a lull between projects or during research phase or just recovering from burnout. There's a lot of very bursty types here though - people who don't do much for a few weeks at a time then drop something massive in a marathon or something.

Isn't an 11% chance to be hiring a really good to incredible employee good signal? I don't think you can have the same chance when hiring from most small companies, not based on company reputation alone.

That assumes that the 11% are on the job market at all.

You're unlikely to see a resume of a distinguished engineer from Google cross your stack of resumes.

For a bit of a blast from the past: https://www.joelonsoftware.com/2006/09/06/finding-great-deve...

> The top 1% of SWEs in any FAANG are world-class engineers and are unique. I'm talking about the John Carmack's of the world who are top of their field.

How are Carmack's SRE/Devops skills?

How are SRE/DevOps skill the measure of a world class software engineer? How good is Vint Serf at SRE/DevOps?

My point is that employers (esp Google/Facebook/Amazon) try to consolidate roles. The term "Full stack developer" encompasses a lot. That's a "10x"/rockstar dev.

Carmack or Cerf probably would not make it through the levelling at Google. They are at a Distingushed engineer level.

Vint Cerf, but point well taken.

Hiring John Carmack to run your SRE team is like buying a Ferrari to only go shopping at Safeway with.

There’s no need to build the right tech on their own. They can simply miss some tech and find a close enough FOSS or build a close enough version.

Where “close enough” means build 80% with 20% effort. Or even 20% solution with a day of effort. Whatever makes sense in the new environment (which doesn’t need to scale to billion active users, for example)

Reply on news.ycombinator.com