I see getting hired as the last stage in the pipeline, with the first stages being your family, then your education and then employment.

The more things go wrong in the first stages, the less likely it is for one to succeed at getting to the top of the next stage.

If people from underprivileged families can only afford to go to the 50% best school in the state, then to the 75% best college, then by the time they enter the job search page, they won't be at the top of the list (again, on average).

I'll use myself as an example: I had the privilege of coding for entire weekends when I was young, while my peers had to work the land. This gave me a distinct advantage early on, which I capitalized on, so at an interview with a company I am more likely to get hired than if those same peers had hypothetically applied, again because it wouldn't make sense for a company to pass on the best candidate when they are all competing for talent, regardless of race, gender, height or other ways of splitting people into groups.

I think everyone agrees that the current situation when it comes to diversity in the workplace is not acceptable, and we should definitely fix it, but in my opinion, if companies are competing fairly for talent, then that part of the pipeline doesn't really need fixing. We can instead use it as a test to see if the attempts at fixing the earlier stages result in better numbers.

P.S. This a difficult subject and I don't intend to offend anyone. Despite my relative privilege above, compared to my peers in the US, I still grew up in a poor and underprivileged family and am behind my current peers both financially and socially, so it's a topic that hits home for me.

How could that be? I was under the impression that a huge amount of new hires come in each year from fresh university graduates.

http://www-2.rotman.utoronto.ca/facbios/file/Whitening%20MS%...

It's a sort of interesting read but nothing too dramatic and whether it's evidence of racism is subject to interpretation. What isn't subject to interpretation is the claim that the study shows that "whitened" names increase callbacks. On the contrary whitening names has no effect on callbacks as indicated on page 31:

"Whitening the name only (versus not whitening at all) did not make a statistically significant difference for black applicants"

The actual study shows that removing racial indicators from experience is what results in a gap in callbacks. So someone who represents themselves as the leader of their campus' "Black Student Business Association" is less likely to get a call back than someone who represented themselves as the leader of their campus' "Student Business Association". They refer to the removal of racial indicators as whitening but I think that's prematurely jumping to conclusions. The name aspect is certainly a form of whitening, since black names are being changed to white names. But it's premature to refer to the removal of racial indicators and making the experience racially neutral as a form of whitening.

Other forms of experience that explicitly mention race result in less callbacks than when that same experience doesn't mention race. Whether this is racism or not can not be concluded strictly based off of the study's parameters. For example, if I were presented with one candidate who was in charge of the university's "Law Society" I would probably pick that person over someone who was in charge of that university's "Black Law Society", and I don't think I would be racist for doing so. I would consider being in charge of an organization that is open to all races or is independent of race as a more prestigious accomplishment than being in charge of an organization that is race specific.

To test whether I was racist in my decision making, I would need to pick someone who was not in charge of anything over someone who was in charge of the "Black Law Society", with all else being equal. The study did not do this comparison and unfortunately the study does not present the raw data so there's no way for me to do this analysis myself.

I could bicker about some of the methodological issues as well which are not exactly rigorous, as well as the fact that this study is not exactly pertinent to this conversation as they only looked at internships for jobs that are not technical in nature, with the bulk of them being sales and marketing, and customer service jobs, instead of engineering, computer science, law, or professional jobs.

Ultimately no study is going to be perfect but one should not read too much into many of these studies. They are not nearly as rigorous or definitive as one would expect and furthermore they don't tend to generalize.

Imagine you having always bought tasty tomatoes from your local farm, and someone steps in to show you studies that determined that the generally tomatoes in the US are tasteless.

I'm also not arguing that tasty tomatoes are taking shelf space away from other fruit.

Given a list of chapters in a video course, and a list of user bookmarks, write a function that groups the bookmarks by chapter. Chapters and bookmarks are defined as millisecond offsets from 0 (the start of the video file).

I'd specify the data structures for junior devs and let the more senior people work those out.

Your average developer that gets hired can solve and unit test the solution in less than 35 minutes. Those who don't might be having an off day, or I didn't do a good job at explaining the problem and details, or they didn't ask the right questions or who knows. The point is, a thousand things can go wrong, but the problem itself isn't that difficult. A few loops and a hashmap gets you an acceptable solution.

This and the other problems in the interview question databases at FAANGs are typical of what's in the Cracking the Coding Interview book. Whether someone's ability to solve these types of problems quickly is indicative of their skill level is a hot topic of debate, but if one wants to join these companies, then hacking the process by learning to solve algorithm and data structure problems is acceptable and not a particularly complex process. It's high overhead for the candidate, but at the very least it proves that they can learn, and learning the custom tools and code quickly is about half the skillet required to be successful at a FAANG.

Which is demonstrably true; they're not hiring from the general population they're hiring largely from universities, and you can look at those universities and see there's already an imbalance in demographics of the available job candidates.