Giulio Starace

Wandering through my mind.

github linkedin twitter instagram email

Rirespira DevBlog 3 | Re-think

Aug 28, 2019
5 minutes read

So, it’s been a while. It’s been more than a month, in fact, since the last time I committed anything to my Rirespira git repository, which includes planning documents.

I’ve been quite busy to say the least; since my last commit I have moved to a new continent and started a 6-month internship which will keep me busy here until the beginning of February.

Moving to a new continent, on my own, where the language spoken is neither Italian nor English, has definitely taken up quite a bit of my time and energy. I’ve had to juggle logistics, bureaucracy, house-hunting and making new friends alongside starting the new job in the last three weeks.

But my move to this foreign land for this opportunity (more about this in another post), is not the sole reason for which there hasn’t been any progress made on Rirespira.

Part of it, I concede, is laziness, I suppose. It’s easy to start side-projects in the summer when there’s nothing going on and time is a commodity. But when your life starts filling up with other tasks, it’s hard to convince yourself to fill your free time with further productivity.

But I suppose one of real reasons why there hasn’t been progress on Rirespira is that I’ve been thinking about it. It and my time in general. Six months, a year even until I supposedly start my Masters is a lot. A lot, sure, but it’s not empty time. Aside from these side-projects which I’d like to work on, I’ll be mostly working on this internship (and probably others), I’ll have to be researching what courses and universities to apply to, preparing application documents, studying for and taking the GRE and it would be vastly wasteful of mine to not take the chance to travel a bit while I’m in this new region of the world.

For whatever reason, I feel a certain pressure to be quite efficient about my time, particularly since I’m allowing myself time to travel as well. Efficiency will inherently be linked to the goals desired; the more goals accomplished with the smaller amount of work and time, the more efficiently spent the time associated with accomplishing them will be. Very simply, I’m ruminating over how I can maximize the number of birds I hit with one or two stones.

So I’ve been reflecting on my goals. Why am I taking up internships? Why do I insist on side-projects? What do I want? I may or may not have mentioned on this blog that I’m quite interested in Machine Learning (wow, shocker, I know), particularly Deep Unsupervised Learning with a focus on Cognitive Machine Learning, how we can learn from our brains to in turn teach machines how to learn. I guess the goal is to embark in a masters so that I can do research on this; I already have a research idea and am excited to get to it. But first I want to improve my skills as a developer. I don’t know whether I want my career to be strictly academic, and I think as such it is important for me to develop more practical skills too, which are relevant to Machine Learning anyway.

The internship I’m working at currently definitely contributes to the improvement of my development skills. There’s a lot of planning, coordinating with people, reading documentation and well, eventually, coding. But it’s missing something, that something very simply being an AI aspect. Now as much as I’m interested in ML and have taken some classes in it, I’d also like to learn more about it and become more proficient at creating projects related to ML before embarking in a masters.

Since my internship kind of leaves a hole in this area, I think this should be filled by my side-projects, along perhaps with Kaggle and autodidactism. So I feel the need now to either abandon Rirespira, focusing on projects more closely related to ML, or readapt it, so that it meets my needs.

I like the idea behind Rirespira. Recently I’ve been feeling more and more inspired to make sure that the work that I do has a directly positive impact on the world, on the things I care about, the environment definitely being one of these. I do also think that in its current form, Rirespira would be very hard to pull off in terms of achieving the popularity necessary for community-powered contribution to be a thing. I never expected Rirespira to take off in any way in the first place and I already have purchased the domain, so I’ve realistically got nothing to lose other than the 7.99$ I paid for it. Rirespira was kind of a side-project for building my planning skills, why not keep utilizing it for experimentation, and who knows, something may come out of it eventually.

So, scratch the community aspect, at least for now. At a very high level, I’m readapting Rirespira so that the raw information on the environmental impact is searched for via web-scraping and parsed, at least initially, via NLP. This is quite ambitious I realize and probably impossible to pull off at the moment. I don’t really know, but that’s kind of the point. This will force me in a phase of research and learning that can only be useful to me. Furthermore, this extends the scope of the project into something larger, comparable to a dissertation, which is something I would’ve liked to do at my university but could only do one in Physics because of silly bureaucratic reasons.

The next steps now most probably begin with the conduction of a relatively thorough literature survey, understanding what the state of the art is in the topics described above and figuring out how these can be implemented.

I hope this will not leave me with a large hiatus once again. As I said I’ve got a quite a bit going on. It’s possible that these are just false hopes I’m giving myself and I’ll end up abandoning the project for good, but I think it’s important to go through these phases of reflections. Either way I’ll have gained something out of this.

See you (hopefully) in the next devblog!



Back to posts


comments powered by Disqus