American dudes getting rich instead of Chinese dudes, though they’ll make the Chinese dudes rich in the process.

Abd while you’re at it, reddit, twitter, any any other SM service I missed.

I don’t think it’s the dogs that have the false positive rate, it’s the handlers.

Sure, there are tons of intangibles that go into it. I’m just saying that people shouldn’t buy because that’s the only way to get ahead, they should buy because that’s the lifestyle they want.

We’re Python on the BE and we don’t really care if they have it on their resume because it’s not hard to learn (that’s why we picked it) and our problems are more complex than just building some CRUD here and there, so we’re willing to teach the right candidate.

So when we do interviews, we let them pick whatever language they want. We’re not testing language knowledge afterall, we’re simply testing logic competence. Can you design a system that will scale when requirements change? That’s what we want to know.

If we use some rules of thumb, it gets closer:

• closing costs - 3-6% of purchase price
• maintenance costs - 1% of current value

Take the down payment and closing costs as an initial investment, the repair costs and any difference in initial mortgage payment vs rent as regular investments, and adjust maintenance and rent (and the difference between mortgage and rent) for inflation. Run those numbers to estimate total wealth after a given period (both house appreciation and investments) and you should end up with pretty similar numbers. I’m ahead on mine as well, but only by ~10% after about 15 years, and my area had really rapid rent growth.

I think it’s an interesting exercise that may not be applicable to everyone since it doesn’t take into account the discipline needed to invest the difference.

I certainly appreciate the honesty, and if we had another role that would suit that person, I would’ve switched the interview to that one instead.

We had someone apply for a FE internship and they were failing, but I noticed they had BE skills in their resume so I switch the interview to that instead. We ended up hiring them for a FT BE role with the promise they could do full-stack if they wanted. They’ve been a great employee since, and I’m glad my boss was able to be flexible on that position (we needed another BE, but hadn’t created the position yet since we needed FE more).

I can’t guarantee everywhere is like that, but I can say interviewing gets old and if we can fill a position (even if it’s not what we were expecting), we’ll do that.

Agreed, but bump the internship up $10-20k for my area. In CS, you should be getting median household income more-or-less for a junior/intern role ($70-80k in my area), and double median income for a senior role ($140-160k in my area), or at least that’s how it works out in my area. There are roles above senior, and those should get paid accordingly.

Yup. We pay six figs (not high six figs, but still double the local median wage) for decent talent. We don’t pay top salaries for our area, but we are about median for a comparable role. The problem is people seem to expect the top roles straight out of college, when they’re really junior level, if they’re competent at all.

This was a problem before AI as well. I’ve been in so many interviews where someone w/ a CS degree can’t deliver solutions to even basic problems. I’ll ask them foundational CS theory, and they can only answer if I don’t expect them to apply it. It’s like they studied for the test instead of actually learning the material. Now w/ AI, they can’t even answer those gimme theory questions, much less apply them.

I used to look for github profiles as a “nice to have,” but it’s becoming more and more of a requirement now, because I just can’t trust someone to actually know how to write code unless they’ve contributed to a larger FOSS project or built something themselves. I can usually tell when they’ve vibe-coded something, so I can disregard most of the nonsense applicants. Unfortunately, this makes it harder for people w/o copious time to get interviews, which sucks (I’ve been there).

Have you done the math on it though? Yeah, a mortgage stays constant, but to get a mortgage, you need a down payment, closing costs, and whatever you’re paying your real estate agent. And then there’s maintenance costs, utilities (most purchasable homes are larger than what you’d otherwise rent), probably extra costs to get around, etc.

If you instead took that down-payment and additional costs and invested it in a diversified stock portfolio, how would they compare?

I’m in a similar boat where my mortgage is now less than half of what rent would be, but my house is growing in value far slower than stocks. Here’s a nerdy video discussing rent vs buy, and the result is that it’s more of a wash than most people assume. This is extra true if you properly account for repair costs (i.e. if you DIY, what’s the value of your time?). The decision to rent vs buy is far less consequential in terms of long-term financial impact than most people assume.

On the flipside, we had a senior position open for something like 2 years before we finally got someone we’re happy with. We fired two before we found someone would would actually do work and not cuss out our external partners.

We’re still having trouble hiring mid-levels. Most of the candidates are surprised when we ask them questions about React when the job position clearly states it’s for React, and they’re also surprised when we ask them to write a few lines of code in an interview (nothing crazy, should take a competent dev 5-10 min, and a nervous one 15, so we allocate 20 min). I don’t think our expectations are unreasonable, here’s how we delineate between tiers:

• junior - needs help from a mentor to deliver feature work
• mid-level - needs direction on larger features, but can deliver independently
• senior - manages larger features, consults w/ architect on high-level design considerations

But all the senior applicants are mid-level at best, mid-level applicants are recent college grads, and junior applicants just finished a coding bootcamp and think they’re hot stuff because they built a rails app by following step-by-step instructions.

We’re not a flashy tech company, we manufacture niche products for a niche field, and our software does simulations and reports. It’s a complicated product, and we’re totally willing to train people, we just want people who can demonstrate that they can ask proper questions and translate that into easy to understand code. The interview questions aren’t hard, but they are intentionally incomplete because we’re not testing coding ability but instead the ability to recognize vagueness and ask clarifying questions (i.e. ask before you assume).

We’re not anyone’s top pick, but we do have a lot of interesting problems to solve and people tend to really like it here. So the candidates we tend to get are desperate people who aren’t getting bites at the flashier companies, which often means they’re not all that competent. During COVID, we’d get maybe 5 applicants for a role after it has been open for a month, and now we’re getting 200-300 in the first few days of the position being open. A lot of those applicants are incompetent and I’m surprised they were offered their previous role, but there are some diamonds in the rough.

That’s ridiculous.

I’m glad my company doesn’t mention specifics. We mention some details about our stack, but we’ve hired people with almost no experience in anything in our stack and we’ve been happy. We want competent professionals who are interested in learning our stack.

At least in our university, web dev was an elective, not a required core CS class. It’s totally reasonable for them to not know how to deal w/ a web server when all they’ve done up to that point is algorithms.

We had a Ph.D work for us who struggled w/ that type of thing. They were absolutely brilliant in their niche (complex 3D modeling of fluid simulations), but integrating their work into our web stack was a nightmare for them (but fairly trivial for us). I asked them to structure their code in a way that would be easy for us to plug in to our web stack, and they looked at me like I was speaking Latin, when all I wanted was a simple entry-point with clearly defined inputs (give me a function to call that doesn’t need a bunch of magic numbers).

If you want a web dev, hire someone w/ web dev experience or be willing to teach them. Not everyone in CS has that experience.

I had a candidate apologize about 10 min in when it was clear that we expected them to know how to actually write code…

I don’t know much about IT (I’m CS), but Ad is very specific to Windows, no? Shouldn’t you instead be asking them about higher-level concepts like access controls, networking principles (http vs tcp vs ip, routing, dns, etc), and basics of cybersecurity (attack surface, network segmentation, etc)? It sounds like you’re looking for practical knowledge about specific solutions rather than familiarity with concepts.

in CS, I can’t expect someone to know our specific stack, but I can expect them to know foundational principles, like data structures, algorithmic complexity, design patterns, design principles, etc. So unless they express familiarity w/ our stack, I keep the questions theoretical, and even if they are familiar w/ our stack, I still keep the questions high level (i.e. for Python, I’ll ask “gotchas” like what’s the difference between a list comprehension and traditional iteration, how does Python’s threading work, etc). I expect them to need to learn something in the first month or two, but also to largely learn on their own. Learning our stack when you’re comfortable with programming in general isn’t all that difficult, learning our stack when you struggle with basic concepts will be a challenge though.

Exactly!

And that’s why I generally ask about FOSS work. If you’re contributing to upstream projects as a hobby, then you’ve demonstrated that you can jump into a larger codebase and figure out their procedures and style guidelines, which means you can probably do the same here.

Failing that, I ask them to apply the theory they learned in school to practical problems, like “how would you use design pattern A for problem X? What about pattern B? How do you decide between the two?” Most people can’t tell me what A or B is, and they can’t even solve problem X with their own methods… I don’t care about people knowing arbitrary design patterns, I care that they can reason about problems, consider multiple approaches, and decide between those approaches given the larger context of the project.

So many people just fall flat on their face in an interview on concepts they should have learned in their third year, which even our people who didn’t go to college can do since they’ve been on the job for a few years. Show me you’re better than a self-taught person and a few years of experience if you want anything other than an intern role.

And this is why I’m not worried about my job, but I do recommend people stay away unless they really like it. I’ve interviewed far too many people who just can’t hack it, yet they have multiple relevant jobs on their resume.

Yeah, I don’t watch him often, and when I do, I’m not a fan. But I’ll absolutely do whatever I can to express how much I dislike him being cancelled.

And there are studies about just that. However, when you review your notes, it matters a lot less what method you used to create the notes.

If you’re unlikely to actually study your notes, handwriting is better. If you’re likely to study them, use whatever is most convenient for studying.