Bloomberg recently published an article saying ChatGPT has racial bias when evaluating resumes. We re-ran their numbers and saw that they didn't do any statistical significance testing. In other words, there's no evidence of racial bias, at least not in this data set. However, ChatGPT has another kind of bias -- it drastically overestimates the value of top companies and top schools and does non-traditional candidates a disservice as a result.
I’ve been an engineering manager at Amazon, Meta and Microsoft. I’ve been in many promotion reviews, and here is my advice on what it takes to get a promotion.
The first super-important thing I learned is that just because you're qualified for the job doesn't mean you can pass the Big Tech Interview. I'm really smart. I got a 4.9 GPA at MIT. By this time I had worked with a pretty wide range of different software languages, tools, frameworks, ideologies, etc. I had managed 30+ people and could also plan and build complex systems solo when needed. I felt very qualified for the positions I was applying for. On my first Big Tech Interview, I did terribly.
Recently, while helping our users negotiate, we've observed a string of aggressive, candidate-unfriendly hiring practices at Meta. We’ve seen the same practices with enough candidates, and across enough different titles and positions, that it’s become clear that they are not isolated incidents or a rogue recruiter’s doing but rather a consistent implementation of a deliberate strategy that comes from the top.
If you’re negotiating with Meta, you need to know how they operate and understand the unwritten rules of the game. If you do not know the rules, you will fail — long before you even start negotiating.
Software engineering jobs come with a lot of perks. But that doesn’t mean our jobs are perfect. Work requires us to commute, reduces the time we can spend with family, increases our stress levels, and forces us to deal with teammates we don't gel with. And sometimes we work for companies with questionable morals and use technologies we don’t enjoy.
For some of us, it’s worth trading cash for a job that fits into our lives better. But how much cash?
To figure this out, we surveyed our users about times they took jobs with lower compensation, why they did it, and how much money they left on the table. We have the numbers!
Does ChatGPT make it easy to cheat in technical interviews? To find out, we ran an experiment where we instructed interviewees on our platform to use ChatGPT in their interviews, unbeknownst to their interviewers. The results were surprising, but as a preview, know this: companies need to change the types of interview questions they are asking—immediately.
Predictions are hard, and, inevitably, most of them turn out wrong. But we’d like to brave the scathing mockery of the internets and try anyway! Our courage is bolstered by some useful data we have (both proprietary and gathered from the internet), which we’ll use to guess what will happen in 2024 and to answer the question foremost in many of our minds: When is hiring coming back?
At interviewing.io, we’ve coached hundreds of people through salary negotiation. We’re good at it — our average user gets $50k more in cash, and we have a 94% success rate.
Having done this a lot, we’ve seen our users make the same two mistakes, over and over, BEFORE they start working with us. These mistakes are costly and make it harder for us to do our jobs. Both involve how you talk to recruiters at the start of your job search, way before there’s an offer. Even if you never use our service, you should absolutely avoid these mistakes.
My name is Kevin. I am not and have never been a software engineer. I have never written or tested a single line of code, and I have never even worked as a PM. Despite that, I was able to pass a Google system design interview.
I had just finished working on a system design interview guide and learned a LOT from doing that, but I also learned a few tricks along the way. If these tricks helped me pass, then imagine what you’ll be able to do with them.
The recent exciting and somewhat horrifying inflection point in AI capability tipped me into writing this blog post.
I simply don't believe that AI can do hiring. My argument isn't about bias (though bias is a real problem) or that it's technologically impossible. It's just that the training data simply isn't available.
Most people believe that if you can somehow combine what's available on LinkedIn, GitHub, and the social graph (who follows whom on Twitter etc.), you'll be able to find the good engineers who are actively looking and also figure out what they want. This is wrong. None of those 3 sources are actually useful. At the end of the day, you can’t use AI for hiring if you don’t have the data. And if you have the data, then you don’t strictly need AI.
Interview prep and job hunting are chaos and pain. We can help. Really.