Somehow it is week 11/12 of my time at Makers. I have no idea where the time as gone, and reading back over these past posts has been a wonderful remainder of just how far I have come on this developer journey, but also how far I have come as a student once again.
The difference between this period of my life, and my former studying experience (as a green 18 year old whilst studying for my BSc degree) is huge. I enjoyed my degree, and my time at university massively, but the difference in my current studies is that I genuinely want to know as much as possible about what I am doing. This curiosity has led to a strange desire to know things that I never knew existed, and to explore areas that I never considered before.
This leads me on to the Final Project here at makers. The final 2 weeks is spent working on a project, no holds barred. You select areas of programming you are interested in, brainstorm ideas, and then through a quick Monte-Carlo Simulation, hope to put in your choice of groups. There are no parameters, no reigns, the programming world is our oyster. This made for an interesting choice on my part. Do I stick safe, to something I am comfortable in, or do i take the plunge and roll the dice?
A good friend in my cohort has a deep passion for machine learning. Over the last few weeks, we have spent many lunchtimes discussing this topic, and all the related areas. As I mentioned earlier in this post, the current rabbitholes I seem to fall into all fall under this umbrella of machine learning, in-particular, those relating to linear algebra and its role in coding. I was always good at math, but this kind of math I have never explored, and venturing into conversations with my classmates, as well as many late nights on youtube videos helped explore this further
So in true Makers style, I decided to go all in, and venture into the 'uncomfortable zone' even further. I picked, and was graced with being selected for the machine learning group. 5 of us will embark on 2 common interests. The first is to build our own neural network that will hopefully play and learn our second common interest - table tennis (int he form of the classic video game pong).
I have spent my fair share of time playing table tennis at Makers, i have found it a great way to unwind and reset my brain, and it really helps with coding. This love for the game led to us joking about a table tennis based final project. Then one night on the internet, I found many articles on a neural net (based on neurons in the body, my old haunt from previous studies) and how they could learn when exposed to a stimulus to create a successful response. Then we got thinking, could we make an AI that learns how to be good at pong?
I remember reading in the news a few years ago that the DeepMind project created an AI so smart, that it could beat the ATARI game library perfectly. Could we replicate this on a smaller scale in two weeks? I could go into massive detail about what we are doing, but quite frankly i don’t have the time, or brain power to explain it all now. All i can say is that one week in, it is like i have a new perspective of coding. What we as a team have been doing this week, compared to what I was doing 3 months ago is simply mind boggling. The world of Python, Django and React are all new and wonderful, and using the scaffolding of the Makers philosophy, i approached these frameworks/languages and made leaps on what I can now do.
I will get around to writing a more in-depth blog about the specifics of the project after we have completed next week, until then many late nights will follow, and more reading into this new and wonderful would of Machine Learning.
As always, and more than ever before, please wish me luck!