DR. SAMEER MASKEY
EXEMPLIFYING THE ENTIRE CONCEPT OF PURSUING UNCONVENTIONAL CAREERS, DR. SAMEER MASKEY IS A NEPALESE FULFILLING HIS DREAMS TO DEVELOP AND GLOBALLY ESTABLISH ARTIFICIAL INTELLIGENCE (AI) BASED CONVERSATIONAL SYSTEMS THAT ALLOW HUMANS TO COMMUNICATE WITH MACHINES.
It is not polite to measure success by numbers and name-dropping is also frowned upon, but sometimes you have to let the statistics speak for themselves. More than 20 papers published in International Conferences and Journals along with 9 pending/granted patents under one’s name is something to worth bragging about.
Dr. Sameer Maskey currently teaches at Columbia University as an Adj. Asst. Professor. His course is designed to bring MBA students and Computer Science students in the same classroom with the aim of bridging the gap between Computer Science and Business management. Case in point: the booming industry of startups that has taken the western world by storm.
But teaching is not his main forte; his expertise surpasses most common occupations. If any character of Nepali origin were to ever be featured in a sci-fi flick, it would probably be based on the works and life of Dr. Sameer. We’d confess to exaggerating, but we’re not particularly sure that we are. He is the founder and CEO of FuseMachines Inc, a company that delves in advanced artificial intelligence which builds software robots that allows humans and machines to communicate in a natural language. To put things into perspective, Dr. Sameer would be a key contributor to a future where you could communicate with robots like you do with humans. Now, the possibilities of that happening are still debatable, but it hasn’t been ruled out either.
More recently, he has expanded his business back to Nepal and has started a similar company here.
His success is one of the many inspirations for people who are contemplating to rebel to the cultural norm of fitting a certain occupation that has been supposed an idle profession. Dr. Sameer shares his experiences and views regarding the issue of choosing off beat careers and choosing to become more than a doctor, scientist or business man. Here’s an insight from a man who dared to break the convention and succeeded.
TNM: Could you tell us, in layman’s terms, what you actually do?
SM: What I basically do is, develop an Artificial Intelligence based conversational system, allowing humans to communicate with machines. It has three aspects to it. Firstly, there is Natural Language Processing. Humans are well adept at perceiving what another person is trying to say based on their tone, expressions, situation etc. It comes naturally to us but it’s a completely different story for machines.
Second, comes the Dialogue Component. To hold out a conversation, each party involved in it have to take turns and have to know when to stop so that the other can reply or respond. This has to be programmed into the machines.
Lastly, there is the Natural Language Generation. There are countless combinations you can make with just a handful of words to mean different things. Once again, humans can do that easily, but it’s difficult to program machines to do the same.
What it does is it understands what you said, processes it and converts it into text and speaks back to you.
TNM: Would you say your program is more advanced than Siri?
SM: I wouldn’t say that. But yes, Siri answers questions from a very large domain while our program focuses on a particular domain which allows the conversation or answers to boast of more expertise in a particular field.
Also, Siri is a voice based application which instigates a lot of errors. Our program is focused on a text to text dialogue which has fewer errors.
TNM: Has your program been practically implied?
SM: The text to text dialogue system has been implemented in NYC Business Acceleration to help people in New York opening eating and drinking (food and beverage) establishments by navigating City agency processes and reducing the time needed to open. This is done via a chat wizard.
The system can work on human mode where the person on the other side of the wizard is a person or on machine/hybrid mode. This is when the questions asked are answered either solely by the machine or by a machine assisted by a human.
Best thing about it is, it is self learning. It calibrates according to the input that people put into the wizard in the form of questions. When a question is asked which can’t be answered correctly, it provides you with options that are closest in proximity to the appropriate answers. The chosen option is calibrated into the system each time people choose it. As the probability distribution for that option increases, it will assume that to be the correct answer after enough people choose it.
Basically, the program learns as more people ask questions. If the machine is absolutely not able to answer a particular question, that question is transferred to a human at a later time. The human operator will answer the question and the program then calibrates it into the system so that the next time it is asked a similar question, it can provide a proper answer.
IF ANY CHARACTER OF NEPALI ORIGIN WERE TO EVER BE FEATURED IN A SCI-FI FLICK, IT WOULD PROBABLY BE BASED ON THE WORKS AND LIFE OF DR. SAMEER.
TNM: Where do you see this going in the future?
SM: To be completely honest, the vision is pretty farfetched. Imagine calling any company in the world and talking to a machine or robot for customer service which answers all your queries as well as a human. If machines can replace the human function, it increases efficiency and that manpower can be better utilized.
We are trying to solve one of the hardest research problems, which is the dialogue system. This involves you talking to machines. If we can solve this problem with good accuracy, there is a possibility of robots that perfectly understand you and you can actually talk to them. Maybe even at some point, machines can communicate with each other.
It’s almost like a sci-fi movie when you think about it.
TNM: You are knowledgeable and you have the degrees to prove it. How were you as a kid? Were you always amongst the smart kids?
SM: I studied at St. Xaviers and I was looking through my year book recently and I came across a part where I had mentioned that I wanted to become a scientist when I grew up. So, yeah, you could say I was a good student academically.
TNM: How did you shift to the business side of things from a scientist?
SM: I started out as wanting to become a scientist, became a scientist, taught as a scientist, did a lot of research and I took all that knowledge and tried to build a business out of it. The good thing about starting a business in the US is you can actually reap the rewards.
It’s called the hockey stick growth, where progress starts off slow but once it catches on, it shoots right up. Considering that the business has the potential. That’s the kind of growth venture capitalists look for.
TNM: What are the prospects of pursuing an offbeat career?
SM: After high school, I enrolled into Army College in Bhaktapur where it wasn’t really common for students to apply for SATs or even apply to study abroad as a whole. In army school you aimed to go into the army.
I applied to a bunch of colleges where I got a number of full scholarships and chose to study in Bates where I started getting interested in computer science. But I had told my parents I was going to study bio medical engineering because back then it was kind of the norm for the studious youngsters to get into engineering or the medical field.
But Computer Science was what I loved and in the end, it is turning out to be a really good choice. The bottom line is, you should always do what you work. Whatever field you choose, it doesn’t matter. When you’re doing something you love, you will be so into it that you will be doing it so much and cumulatively the payoffs will also definitely be seen in the long run.
IF WE CAN SOLVE THIS PROBLEM WITH GOOD ACCURACY, THERE IS A POSSIBILITY OF ROBOTS THAT PERFECTLY UNDERSTAND YOU AND YOU CAN ACTUALLY TALK TO THEM.
TNM: Is Nepal a good place for start ups, practically? Or is it a better idea to initiate startups abroad?
SM: The good thing about startups in Nepal, especially tech startups, is that they don’t face a lot of competition. While in places abroad, like the Silicon Valley, the competition is intense. Everyone wants to do a start-up. However, there is a lot of money involved for startups from venture capitalists and other investors in the US, but Nepal still hasn’t developed an environment where startups can thrive.
So even if a revolutionary idea does come up, say an app, it’s just an app and not a business. To start a business there should be prospects for investment, and there aren’t enough investors willing to take the plunge here.
In most cases, you only hear about startups that succeeded. But the fact of the matter is, 9 out of 10 startups don’t make it big. Nepal might not be ready to play on such odds and take such risks with big investments. Furthermore, even if there are only 3-4 startups created in Nepal, the probability of even one succeeding is low.
However, the lack of competition can be an advantage for people aiming to invest in startups.
TNM: From your experience, how important is commitment and dedication for success?
SM: If you have a startup, or any other business for that matter, you have to live it for it to be successful. You need to be doing it all the time. You have to put some skin into it, you have to dedicate yourself.
TNM: Do you have any suggestions for people choosing off beat careers?
SM: There are three things you have to keep in mind, in any career path. First you have to love your work and to do that you have to do what you love. It does not matter what it is. Secondly, you have to be proactive. Go and seek out opportunities, no one is going to bring anything to you in a plate. Lastly, you have to dedicate yourself to whatever you are pursuing.
PHOTO: NIREN TULADHAR