This snippet interview is from interview 13 of the Bangladeshi Perspective with Adib Hasan. The full interview is available on our page in Facebook.
A bit about Adib Hasan:
Adib Hasan has won three bronze medals at the International Mathematical Olympiad. He is currently pursuing a double major in Computer Science and Mathematics at the Massachusetts Institute of Technology. Professionally, he is interested in designing machine learning augmented algorithms. He is also a member of MIT’s Machine Intelligence Community. He loves combinatorics and has co-authored a book titled “An Introduction to Combinatorics (Part 1, Part 2)”, which is one of the first books to be ever written in LaTeXbangla, a package that Adib created that simplifies the process of writing Bangla documents in LaTeX.
You can find all the other interviews above. Continue onward to read the interview snippet.
Paraphrased Translated Transcript is provided below:
Adib: So, I actually kind of didn’t want to study ‘pure software engineering’ because pure software engineering is, well, it’s my philosophy, which isn’t really that important to be explained. Anyway..
Seeam: What is it? I’m really interested in knowing that.
Adib: Well, there’s a lot about software engineering. I would say like in many cases it’s not creative enough. I mean in many ways it’s not that creative of a thing. You really don’t need to look that deep into it, in the sense that different types of API or libraries already exist and you can make something using them.
And if you start from the very beginning, maybe if you become a senior programmer and if you have a big team under you, then you can take a creative decision like I can do this in so-and-so way.
And if you’re just a programmer at the beginner level, then it’s like a decision has been taken in the meeting and now you have a part that you are to fulfill. So you have very few chances of making creative decisions. I mean you can’t show much creativity.
Seeam: Isn’t that the case in most fields, for people in junior positions?
Adib: Not really. Like I can give an example – Like in case of ‘data engineering’ which is like kind of working with data, the convenience of working with data is that if you have a data set within your reach, something that can be found in the data set can’t be found without it. Because without seeing the data set you actually cannot say what you’ll get. For such this seems very interesting to me. I mean it’s not only for this reason, but I had a liking for it from a long time ago as, in different cases, math can be used in it in many ways.
Well, also in this sense, that almost everybody is doing ‘deep-learning’. And I’m laughing at everyone because one problem with them is that they think they’ve learned this thing and understand ‘machine-learning’ and stuff related to it. But suppose, how do we humans come to know and comprehend how something functions? The present scenario with machine-learning has become such, that you can make a model even without having the sufficient knowledge as there are so many libraries right now.
Okay, you made a model. But for some reason the model is not functioning. In such a case, you won’t be able to find out why it isn’t functioning because you still didn’t understand the simple and basic thing, which is the model’s underlying mechanism.
Well, I can give an example – Majority of people know this that, in many cases, to estimate a probability we use ‘cross-entropy’ philosophy or (negative clock glide?). Now my question is, what is it actually measuring? I see very few people thinking about this thing. I mean, to know is not important but what is important is, has anyone thought about it. They are like – Oh, there is this law and so let’s apply it.
But, why is it functioning and how was it conceptualized? They have very little curiosity regarding this matter. So this is just a simple example.
Seeam: Okay. Yes, I guess all the tutorials available online are similar.
Adib: In the tutorials they tell you that, only if you input the values then you’ll create something magical, as if you will become a magician.
But why is it happening? How is it happening? I mean try to understand the fundamentals, the concept or else your knowledge will be limited to the tutorial only. This is very important. Because to do something on your own, to be creative you need the in-depth or foundation knowledge.
I’m not saying people don’t have this urge in them. Many have it. Not knowing is not a crime, but at least you have the scope to learn new things. What makes me sad is that, many don’t even have the interest to know and comprehend the subject just because they think there is no use in learning the basics as they already have necessary resources to get their work done. It’s like, in deep learning I have this and that does the job, so I don’t need to have any basic understanding of the subject matter.
There is a massive difference between not knowing something and being curious VS not having the curiosity at all.Adib Hasan
This interview was transribed by: Fateen Tahseen Alam
Fateen Tahseen Alam is a 1st year student at the University of Dhaka studying Mass Communication and Journalism, and is passionate about promoting “Empowerment through ideas and skills, Sustainability through empowerment”. He is an ardent supporter of professional development in communities around him. Email him at firstname.lastname@example.org to express opinions, requests and compliments.