tangential thought, but related: I am, in general, a proponent of nuclear energy as a green alternative to whatever the hell we are doing today. But when I see such stories that humans manage to fuck up simple payment processing apps, still make errors while maintaining bridges, still manage to do hugely negligent screw-ups (most likely corrupt) in *cable cars maintenance*, I immediately think that it is imminent, that something will go wrong with such complex thing as a nuclear reactor and the price there is much bigger.
I kind of get the worry, but the requirements and processes seem to scale exponentially with reliability needs. Online companies may fuck up every day in new and creative ways and we barely get to hear about it. On the other hand we know of every nuclear failure so far with enough public details to discuss the whole time line, system design, steps each person followed, etc. and the death count is still minimal. Then each of those is an input to the future processes. Nuclear power plants and air traffic are in their own class of reliability and safety processes - not even comparable to that's happening in internet commerce.
We know every nuclear failure. We don't know every time a strong nuclear risk existed but by chance, didn't trigger. Nuclear power plants are probably much safer on average, but it only takes one corner cutting plant to cause a nuclear accident.
The Italian cable car was really messed up. The emergency brakes of that cart were intermittently triggering, so the operator jammed a piece of metal to stop that from happening.
His assumption is surely, "Relax, what's going to happen, the cable won't break!".
> The emergency brakes of that cart were intermittently triggering
My guess: each time a strand within the cable broke the cable stretched a little and the brake triggered.
Five years ago a company was hired to maintain the cable car. They took one look at the state of it, wrote to the operator (the town council) saying it needed to be shut down and exited the contract. It was an accident waiting to happen long before the brake fiasco.
There are also plenty of services with really incredible uptime. You just don't hear about them because they're deep in whatever stack you're using and haven't broken publicly in decades.
It's all about good engineering practice and architecture.
In fairness, the government does not watch over your shoulder if you write code for payment apps. Nuclear energy oversight is so over-the-top, it's almost not worth doing it.
There are much-safer-by-design reactor plans out there now. Hopefully the "nuclear is horribly unsafe by default" opinions will consider the new options.
I thought that IQ test was screening test, pre-phone interview. But no, they had me redo it at the onsite interview too. The funny part was the onsite test had the exact same questions as pre-phone interview.
I remember doing it too. I was at work in a meeting and they have instructions saying something to the tune of finding a quiet place and all of that, but my thoughts were if they are serious about this, then solving these abstract problems is something I'll have to be able to do while under pressure or under the heat of conversation.
Long story short anyway, I'm not intelligent enough to work there I guess, so good thing they used that test to screen me out and make sure I knew. It does have a little bit of merit with the very quick no versus the long, drawn out no. I recently interviewed at a great company, 4 1-1 interviews, a presentation/demo I had to make to present to 7 other people, etc. and I think another interview after that and I'm just over it.
I've had to take the MMPI [1] for an employer before. About 500 true/false questions to screen for mental health disorders. Some of the questions seem quite outlandish but taken as a whole make sense.
This is exactly why these IQ-test companies make so much money. It gives out yeses and no:s confirmation bias does the rest.
Quickly why they don't work:
You create a huge chain correlational assumptions. First that visual-spatial tasks of this kind predict performance on visual tasks. 2. That performance on visual tasks predict general intelligence (whatever that is). 3. That this notion of general intelligence (which is usually and arbitrarily defined not to include social skills) actually correlates with the tasks that you think the person will be performing, and finally that your idea of what the role has an impact on the company. Of course it is completely absurd, what they are selling is snake oil, plain and simple.
The remedy I recommend is simple, talk to the person - do it and you will be able to tell within 5 minutes.
I use it to filter out candidates when hiring assistants from the Philippines. I get so many applicants that I don't have time to talk to all of them. Have you been on the hiring end? You get hundreds of applications within a couple days when you have a good position and are paying well enough.
There is a strong correlation between IQ and professional achievement whether you want to believe it or not.
Right - and the general intelligence thing is funny too because all of these companies want to hire specialists in some area, not generally intelligent people.
I think you're just trying to justify an unjustifiable test to protect your ego.
First, if you're just hiring assistants from the Philippines they could just have someone else take the test or get around it some other way.
Second, you have no good data to support this hiring practice. You're free to use it, but it's no better than just hiring a random person from your pool of applicants. You might as well screen based on their favorite color too to just make up filtering criteria.
I failed a job application because of the IQ test. It was administered in a second language for me, so I really didn't do well... the interviews had been completely smooth and I got on well with everyone I had talked with...
The thing is, I really needed that job... ended up going to another job that offered me a very low salary (I had no visa in the country , so was looking for a sponsor, which makes things a lot harder) and the company went bankrupt within a few months!
Anyway, I still got the visa, and then, with a few months to find another job with more peace of mind, I eventually got much nicer job, paying a lot more! But I still dread the though of doing an IQ test, despite my years of experience indicating I am more competent than average, at least.
I had to do it twice for different companies that used the same IQ-test platform.
And most of the time it's not even proper IQ-test but only Raven Matrice test + maybe quick math tests.
Funny thing was that I did very good (apparently according to the HR person) on one of them, but did horrible enough they didn't even call back on the second test.
grids my gear why this is still a common practice in Sweden. HR in Sweden seems to be about one or two decades behind rest of the world in their efficiency.
They did it a year ago when I was applying there. I was so annoyed it almost made me cancel the interview. In the end I canceled, because they didn't allow working from home.
Doesn't almost everybody? I mean let's get real here, what's the IQ distribution at FAANG or any other competitive software engineering company?
There may not be an explicit "IQ" portion to the process, or a hard number, but they are absolutely filtering on intelligence. An uncomfortable aspect of our society that I'm both surprised and not surprised doesn't get talked about much.
Yep, took it just a month ago. Quite dumb honestly, not sure what it indicates. It was a bunch of weird pattern matching and guessing? Seemed easy, but got a rejection a week later.
I don't think they do those anymore. At least when I've interviewed there (2x) over the last couple of years I did not encounter any of those types of questions.
It's not an IQ test. It's just pattern recognition which is about 5% of the tasks you do in a real IQ test.
When I joined Klarna in 2011, the test was so easy that I joked I could score full marks on it even if I was hungover with no sleep. There was one question on the test that actually had 2 correct answers depending on what logic you applied. This was actually a real issue when recruiting, because there was a hard cut-off to make it into the engineering department, and several times I had to ask "what was their answer on question 12?"
It caused quite a bit of commotion at HR to change the official test scoring to 2 correct answers for that question.
Now the test is like a million times harder and your score at the end is between 0-10 and you have no idea how many questions you actually answered correctly. I would be very interested to know the "true" answers of these new tests to understand what kind of crazy logic you need to apply to get every question right. I'm almost certain it would take me longer to understand the answer than the time you have to do the test.
That test was always stupid and fought hard by a lot of engineers that considered it so. It was still kept even after a lot of pushback. I left the interviewing team due to that, I couldn't be part of a process that considered that step not only required but as a hard cutoff for engineers.
I lost so many great candidates that would be great hires to my teams at Klarna to that stupid test.
