4 ms·
TL;DR: The barchart can be misleading. Increasing # of applications can decrease the rate at which they mature into offers. I like some of your points like get
by goldenbeet 10y ago
TL;DR: The barchart can be misleading. Increasing # of applications can decrease the rate at which they mature into offers.
I like some of your points like get through to real people. That's an incredibly important part of landing a job. However I do have an issue with the graphic at the beginning. I think it's misleading, because you shouldn't really be applying to 200+ companies. When you take that kind of approach its like you're selling yourself short (not putting your best foot forward). Its like saying "I'll take anything, just please give me a job". It's also more difficult to actually show interest in whatever company you're talking to, which is a key part of performing well and ultimately obtaining an offer. So you end up showing 8 offers from 291 applications, but I would argue that by making so many applications, you are decreasing the rate at which the applications turn into offers. I can offer a contrasting viewpoint to this job search thing. When I was in school for Physics I taught myself how to program and started focusing on Web Dev. I ended up dropping out of school and decided to become a programmer. In that initial job search I applied to three companies in SF. (Companies that I was familiar with and legitimately interested in) and got offers from all three. That was two years ago and I've since applied to two other companies in Mountain View (I'm trying to move to an ML engineer position) and received offers from both as well. And like I said I think the difference between what our bar graphs would look like is due to the fact that my applications were much more targeted and my interest and motivation in the companies was much more apparent.
- blunte 10y agoI'm not surprised by this, neither by the methods and approach he employed nor by the results and education he received from his efforts. The fallacy that "programming" is just learning tools and syntax and processes has unsurprisingly led to an expectation that, in continuing fashion, one can just run through a defined routine and get a solution. In short, the whole thing is misunderstood, and thus the approach is all wrong. Whether in IT or business (or life in general), opportunity is more easily gained through human interaction - networking (but substantive, not superficial). Go to meetups on a particular language or stack or tool. Become involved in open source projects or communities that are centered around some shared interest. Get to know people, and let them see who you are. Then the jobs will come to you.
- votr 10y agoWhat was your strategy for dealing with how most ML positions require advanced degrees?
- goldenbeet 10y agoSo I was realistic about the companies I could apply to with no PhD/Masters (No Facebook AI Research for me). I also made sure I could have some sort of tangible projects completed. (Created a GAN to generate images of street signs, DL to predict steering angle for a car based on video feed, etc) Then I went around to meetups, workshops, conferences to meet people and learn. Then through talking to people in person I found some smaller companies that are implementing ML in interesting ways (self driving cars, building recommender systems) and applied to the two that I felt strongest about. The key though is that I met people from the company and talked to them about what they're working on and what I've been working on and if there's overlap, it shows that I can contribute to what they're working on. Every position has a requirement of some sort of degree, but I just ignore that and apply anyway if I've already met and talked to people at the company.
- votr 10y agoGood tips, thank you. I started in algorithmic trading with Java and Entprise-y stuff. Got bored and moved to front-end. Now looking to move to ML. Like you, no degree, but I apply anyway as long as I feel I have the right signaling.