Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
amznthrowaway5
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
14 ms
·
91.
▲
by
amznthrowaway5
6y ago
I wouldn't consider cutthroat deceit "professional" behavior, but that's the depressing reality of how the world works.
92.
▲
by
amznthrowaway5
6y ago
Yes, those misaligned incentives mean rewarding the cutthroat psychopaths who are obsessed with claiming credit for others work, instead of genuine team players who care about delivering a quality product. There is lots of cronyism, like al
93.
▲
by
amznthrowaway5
6y ago
Not true in large companies. Managers are often just empire building, but new hires still get top of band. Real morale killer, makes tenured employees want to look for a new job.
94.
▲
by
amznthrowaway5
6y ago
At least for engineering manager positions, there shouldn't be any non-technical people. How are non-technical managers supposed to allocate work and judge how well their technical employees are performing.
95.
▲
by
amznthrowaway5
6y ago
In many complex domains like ML, I find most people don't contribute anything and actively make the problem worse. You often don't need the amount of people that large companies often hire, it's just empire building. Software
96.
▲
by
amznthrowaway5
6y ago
Jack Welch style stack ranking and forced attrition of bottom 5-10+% yearly is also one of the "benefits".
97.
▲
by
amznthrowaway5
6y ago
The "signing bonus" from Amazon isn't really a signing bonus, it's just to make up for the fact that they are not giving you much stock for the first couple of years.
98.
▲
by
amznthrowaway5
6y ago
Most people at FAANG may not be very sharp, but very sharp people easily make it into FAANG and it's a much better deal than the majority of startups.
99.
▲
by
amznthrowaway5
6y ago
> you'll get word that somebody spun up a working group to investigate X That "somebody" is often one of the yes-men of the manager, these are the types of people that management usually rewards so of course they are going
100.
▲
by
amznthrowaway5
6y ago
That's not a good parallel at all, the manager still has the real power and can fire/PIP the employee for making that joke.
101.
▲
by
amznthrowaway5
6y ago
These policies and processes like PIPs are for protecting the company and managers, not the managed employees. Most of the stuff on these PIP documents are usually just for leaving a paper trail and avoiding lawsuits, anything can be exagge
102.
▲
by
amznthrowaway5
6y ago
Since the estimates are almost always off, it's a perfect way for the managers to reward the people they like for political reasons, and punish the ones they don't.
103.
▲
by
amznthrowaway5
6y ago
> That sounds lovely in theory but gets tricky quick with employees who give you estimates that seem too high, and then don't even meet those. How many managers are competent enough to judge that kind of stuff, and whether or not an
104.
▲
by
amznthrowaway5
6y ago
This is also the experience I've had with PIPs, "performance" reviews and promotions at large companies. Also incompetent but well connected people are some of the most likely to be promoted into management.
105.
▲
by
amznthrowaway5
6y ago
You really shouldn't, those people are likely corrupt as well and may get you deported.
106.
▲
by
amznthrowaway5
6y ago
Unfortunately the data science craze has resulted in a lot of these corrupt academic politician types coming into industry, so you definitely don't want to go into "machine learning" if you want to see decency.
107.
▲
by
amznthrowaway5
6y ago
Yes, once the corporation gets large enough it becomes almost purely political.
108.
▲
by
amznthrowaway5
6y ago
At amazon a lot of the "engineering" managers are so incompetent they can't write for loops.
109.
▲
by
amznthrowaway5
6y ago
Here it is, from peopleware: > I was teaching an in-house design course some years ago, when one of the upper managers buttonholed me to request that I assess some of the people in the course (his project staff). He was particularly curi
110.
▲
by
amznthrowaway5
6y ago
Exactly, there is still a lot of corrupt credential signaling in the tech industry, but tech is one of the best due to leetcode interviews.
111.
▲
by
amznthrowaway5
6y ago
> People dismiss college credentials as elitist gatekeeping, When you create entire artificial domains out of whole cloth, like "competitive programming", that's gatekeeping on steroids. No it isn't. Anyone can pick u
112.
▲
by
amznthrowaway5
6y ago
> But then - why do anything more than the bare minimum at your job to get by while devoting the rest of your time (including any downtime during your job's working hours) to study leetcode for interviews instead? At least until you
113.
▲
by
amznthrowaway5
6y ago
> I'm not surprised that Siri gets things like this wrong, unlike every traditional search engine. Each text based search engine returns more than one possible response, and if they return the wrong answer first, they can monitor ho
114.
▲
by
amznthrowaway5
6y ago
Greatly misleading or lying by omission isn’t much better than lying directly.
115.
▲
by
amznthrowaway5
6y ago
> It helps to reframe the issue. "Selling" yourself is distasteful for many, because it smacks not only of loudness & brashness, but of straight up lying. But a lot of people who get the promotions and funding are straight
116.
▲
by
amznthrowaway5
6y ago
In my experience working on these product teams, most of the real gains are just from solid engineering combined with public advances in ML. The teams filled with ML specific folk aren't able to accomplish much.
117.
▲
by
amznthrowaway5
6y ago
That's true in more than just in the self-driving car space. Even at companies like AAPL and AMZN where ML focused researchers/engineers work on production related tasks, I've seen their value production is dubious at best.
118.
▲
by
amznthrowaway5
6y ago
> Competence is very far down this equation. The worst part is when performance review time comes around. It's obvious to anyone looking at the situation objectively that competence doesn't matter much when choosing who to rewa
119.
▲
by
amznthrowaway5
6y ago
> The Role of a Data Engineer on a Team is Complementary and Defined By The Tasks That Others Don’t (Want To) Do (Well) Those types of tasks are also often not recognized or rewarded by management, despite being a hugely critical part of
120.
▲
by
amznthrowaway5
6y ago
Good designers can't design without a deep understanding of the engineering. Non-technical people shouldn't be driving or designing engineering projects.
More ›