A reflection on my master's course

After getting my Bachelor's degree, I immediately got a job working in the corporate world. After a few years grinding away at a tedious job, I figured it was as good a time as any to refresh my skills. 12 months ago, I made the risky decision to leave my job and come to Canberra to do my Master of Computing at ANU, specializing in Artificial Intelligence.

So far, I only regret one thing - I should have done this sooner.

Artificial intelligence is a vast and exciting field. We are now getting at a point where the AIs of science fiction could be reality within our lifetimes. We already have phones that can talk to us almost like a human will. The last time a human beat a computer at chess was in 2002. States have to draft legislation to regulate self-driving vehicles, because they are no longer hypothetical.

It was the possibility of being able to contribute to this amazing field that drove my decision to pursue the Master's degree. Now that I am more than halfway through my course, here's a couple of thoughts I have about the experience.

Self-learning

One decision I grappled with early on was whether to pursue formal education at all. Nowadays, there's something called a "Massively Online Open Course", or a MOOC as it is clumsily called. MOOCs are essentially online courses, designed so that students can learn on at their own pace, in the comfort of their own home.

The latter option was attractive - it would have meant I don't need to quit my job and move to another country. Most MOOCs are also free, which is a huge plus. Before coming to ANU, I actually completed a course on Machine Learning on Coursera. I highly recommend it.

However, I ultimately decided that the certification, and the access to some of the smartest minds in the field, was worth the extra cost. So far, that decision seems to be paying off.

Doing the Master's degree had the nice side-benefit of allowing me to meet with other people with an equally strong interest in this field, something which is lacking in Malaysia unfortunately.

Maths is hard

Any first-year undergraduate CS student can tell you that an AI is nothing more than just a slightly more complicated piece of software. When I first started this course, I came in thinking that 4 years of programming out in the real world should be more than enough experience.

Boy, was I ever so wrong. Never trust first-year undergrads.

The vastness of the AI space makes it a tough field to get into. It draws heavily from statistics, probability theory, linear algebra, and even a little bit of linguistics and psychology. This meant that you need a very strong grasp of foundational maths to be able to keep up with the coursework. Something which I didn't realize I lacked until I started the course.

One of my biggest weakness was that I allowed my maths skills to rust over the years. This is ironic because people who know me know that I used to be borderline obsessed with maths as a kid. I was probably one of like 3 people from my high school who genuinely thought maths was interesting.

However, what happened is that somewhere along the line, I chose to believe that maths was useless. Nothing more than an idle curiosity that has no applications out in the real world. I stashed the maths in the attic of my mind, and decided to start working on actual problems. Like making a website look pretty.

Thanks to the Master's course, I know now that the "maths is useless" mantra is a wrong one. All the cool stuff I mentioned earlier, they are built on top of some fairly complicated mathematics. If you want to build a voice-recognition software which allows you to talk to computers, you need to understand Hidden Markov Models, which in turn requires a solid understanding of probability theory. If you have a strong understanding of statistical inference, you could build Google Translate. Formal mathematics is a useful skill to have, and it's not something you can get working in the private sector.

Conclusion

In short, doing my Master's degree has been a fruitful endeavour for me so far. "It's all about the journey, not the destination", as the old cliché goes, and this journey has been an excellent one.