What's a good thing to learn right now to prepare me for the future?

2020-01-15 Nick Larsen

Dear Nick

I'm a Computer Science student currently in the middle of the course.

I've been thinking a lot about my future career and, since computer science is a really broad field, I have to say that I'm a little confused about which path should I pursue.

Everywhere I'm seeing that machine/deep learning is a growing field, with lots of good opportunities, and in my university there are a lot of research being made in the field. One of my professors called me to participate in one, so, that's the path that I'm thinking about right now.

One of the concerns that I have is the fact that that are so many people studying the field, I'm always thinking: "Maybe in a near future there will be so many AI people and not as many jobs out there...". Could you give me your opinion as someone with a lot of experience on the industry? I would like to know from someone more knowledgeable if I'm really making a good choice focusing my effort on it.

Well, having said all of this, my objective trying to find someone to mentor me is to have some guidance on my studies. I have no problem studying by myself, but I'm always starting and stopping, maybe because I'm not good at establishing short term objectives.

There's a lot to discuss for sure, but I think that's enough for now.

Best regards, Luan

Dear Luan

I'm happy to help you stay motivated and keep moving on your studies. I love AI and ML and it's awesome that you are interested it in it, but before we go that route, let's talk about your concerns.

Everything is just hype. As you will see time and time again, in not just our industry, but also in just about every industry, people need to be excited about stuff in order to invest in it, so the hype machine gets started and everyone who wants some money to pursue their dreams jumps on that train and perpetuates the vision. Take for example self driving cars; I absolutely love this idea, it solves way more problems than it creates and it has the potential to make the world an enormously safer place to live for everyone. There were significant advances made in this space and it got to the point where even major car manufacturers were declaring they would have consumer ready self driving cars available in 3 years. That was promised as far back as 2016. Obviously that didn't happen. It was just a really good story and the hype train was having a giant party on board. Turns out driving cars is constantly seeing new situations and our existing methods for training models do not handle unseen conditions even good enough to go to market. It's a super hard problem.

The world is full of these hard problems, and while AI and ML seem promising to make advances in these areas, it's nothing more than hype until someone actually does it. Will there be less jobs out there? I'm not so sure about that and these are just my ideas, but it seems unlikely that the jobs will dry up anytime in the next decade barring AGI coming to fruition which it is essentially consensus that wont happen. I think of AI and ML just like I think of sorting and searching, they are just problem solving techniques. You got a problem where you need to locate a user record in a database, you use searching, you need to organize a list from most important to least, you sort it, you need to find directions on a map, you use AI, you need to predict the price of a house, you use ML. It turns out there are a lot more interesting problems that AI and ML can be used for that they haven't been applied to yet. I think in the near future AI and ML knowledge will be required for far more positions. You won't be a data scientist because you know it, you'll still just be a developer, but those kinds of techniques will be necessary to keep your solutions on par with the rest of the industry. It's just a new button on your calculator.

I also see professional certifications becoming a requirement for specific job roles in CS, a lot like lawyers have to pass the bar and how doctors can pursue a board certification. There are a lot of ethics concerns in the problems that we are applying AI and ML to (moreso ML) as well as a lot of common mistakes which invalidate the science aspect of the work and having a professional certification seems like the most natural way for practitioners to convince people of their skills. Will you need something like this for making web apps? No of course not, but for the jobs where you can create an enormous amount of ill, it's long overdue.

Programming to me is just a really fancy calculator. It's not that useful on it's own but it can save you enormous amounts of effort when you use it to help you solve other problems which brings us back to your first question, which field should you pursue? The answer is simple, try them all, choose the one you enjoy the most. Don't worry about your future prospects, your long term happiness depends on enjoying the work you do. Despite what people say, you don't even need to be passionate about the things you do, you just need to feel like your work has purpose and that will be enough. Go find your purpose; try as many things as you can right now and figure out what problems you want to solve, then just try to solve them. When you're focused on solving a problem, you'll learn whatever you need to in order to solve that problem and you won't feel like you're limited to being an ML dev or a web dev or whatever. You'll just constantly learn new techniques to apply to your problem every day until you stop making progress, then you'll go learn some more.

I'm happy to answer any other specific questions you have about the industry, some ideas about problems you might want to work on or I can help you further understand the techniques you are currently learning. Feel free to ask anything.

If you are currently looking for a mentor I'd be happy to help! Please reach out to me by email with your specific questions and I'll get back to you as soon as possible!