This year during Hactoberfest we decided to run a month-long study group in the form a WhatsApp group.
I read the following somewhere, this makes perfect sense to be applied to this study group:
The formula to induce motivation, in my experience, starts with action, which sparks inspiration, which sparks motivation, which leads to action, and the loop continues.
So we decided to have WhatsApp AMAs in the group, which turned out to be a super fine idea. It encourages participants to ask more questions as well they get to learn from the experiences of interesting people.
Last week on 12th October 2019, we invited Ashris Choudhary for an hour-long AMA.
- A: Ashris
- P(n): Participants
After welcoming Ashris to the WhatsApp group, he greeted everyone and introduced himself with a video. The following is a trimmed and slightly reordered transcript of the AMA. If you want the full chatlog you can access it here.
P1: My question is that How do you choose project? How you select the difficulty of the project?
A: So I have done the Hacktoberfest twice. The best way I think is to look for the Hacktoberfest tag on Github - lot of people tag beginner friendly bugs for you to solve, that could be a good start. You can even work with some college seniors on their projects - usually someone who has worked in Hacktoberfest before is a good idea. Look for something that is in your comfort zone, is beginner friendly and can be done using 10-15 lines of code. Discussing with the repo owner about the complexity of the bug you are trying to fix can be a good start. Wish you all the best! this is a good starting point.
P1: What would you suggest -mastering one particular domain or jack of all trade?
A: Definitely master one domain in the long run. Find one domain where you are in the top 5% of the world. Now, the interesting thing is ‘domain’ is not defined well. I believe we are living in a world where domains are merging to become new domains. Like psychology, neurology and AI combine to form cognitive augmentation. The established domains like physics, maths are all saturated, it is extremely hard to shine there. So for the initial years, like now, try to be a jack of all trades - do many things until you find the right mixture, make it a domain and be great at it. For example, I was good in design and code. Was I an awesome coder? Maybe not. Was I an awesome designer? Maybe not. But when I combined them two and did data viz, I unlocked a field not many are doing and it was easy for me to do things. I recommend this route.
P1: How to get into MIT media lab like you?
A: Work on crazy projects, have a stellar portfolio, reach out to professors personally and learn new things every day - even if it doesn’t guarantee Media Lab, you will easily be ahead of 99% of people your age.
P1: Should the projects be something that no one ever tried which is very new?
A: Yes and no - Media Lab values interdisciplinary approach to building things. You can take two normal but unrelated skills and fuse them together to make something new. You don’t have to reinvent the wheel, just find new ways existing things can be paired. Be really good at one thing - electronics, code or fabrication and show how you can use them like artists use colors, poets use poetry to create magic.
P1: How to get into top IT companies ? some people say competitive programming help a lot, some says having interesting project helps. What would you say?
A: Well, each company has its own way of picking people. Once you are being interviewed, it is your acumen that matters there. The tough part is to get an interview after being shortlisted through the series of automated tests. Some hacks help like having the keywords in your resume that companies are looking for. But nothing beats just being plain good in your skills. Some sure shot things that work is to be more visible - go present your work in conferences, take up GSoC, participate in the Facebook bug bountry programs, start helping startups in their open source projects - if you do all this, may be not the top4 [facebook, google, apple, amazon] come to you, but many IT companies will line up for you to join them.
P2: Hello Ashris, Thank You for your time. Media Labs must be awesome experience. I’d like to ask , what are your views on startup teams , Suppose you have a startup and, you’re taking in more members , like what qualities do you look in a person who’d want to join your team.
A: This is such a relevant and a great question. When investors look to find ideas, they look at the team also. Not only should your idea be impactful but you should prove why you are the right people to solve it. At MIT, in one of the entrepreneurship classes, I was told that every good startup needs a combination of 3 kinds of people - the hustler, the hacker and the hipster. Hustlers are the ones who know business, can do sales and marketing. Hackers are your tech wizs and Hipsters are the Design/product wizs who know why to build something. When I try to team with someone, I see which domain do they belong to and how much of self driven projects have they done - because that shows that they are genuinely interested and skilled in what they do.
P2: Awesome , got it , Hustler - Hacker - Hipster. Also , I’d like to know , what is the stage/state of a startup when it starts getting funded ? Like , (1) the idea state (we can do this that wala state), (2) some tid bid research done wala state , (3) prototype ready wala state (4) some revenue generated wala state.
P2: Also , another question , like what would you prefer , the attitude of a startup leader should be - (1) day /night we have to do this anyhow , some team members have some errand to do ,go later complete now wala attitude , or (2) or go , do your errand and we can complete this later wala attitude. ?
A: Interesting. As a startup founder, you are someone who is creating a new thing in the world. You are like a mother to a baby. Can a mother leave the baby and just chill? No, the baby is too fragile and it will die of negligence. A startup is like that, you have to nurture it continuously. When you have employees, you cannot babysit them, they need to feel part of the product also to give their 200%. So if you are serious about a startup, it should be everything of your life - something you think when you sleep and when you wake up - this is why startups are not for everyone. It is bleed for 5 years so that you dont have to bleed ever again wala mentality.
A: Definitely 4 - someone will pour money on you when you already have set the fire and their money would just help it blaze more and earn more money 💰💰💰. You have to solve a problem of society that society doesn’t yet know how to fix. There is no short cut. No prototype or website is going to fake it. You need to find someone willing to pay for your product, only then can you even think of an investor.
P2: What does it take to make it to an Ivy league Masters or Maybe CSAIL Masters , like is it Research Work done of similar kind to the projects ongoing at CSAIL , or is it something different (like entire portfolio) , how far GRE marks/ institute name holds strong to admission process.
A: Depends on the uni but yes, your institute, your GPA, your GRE, Letters of recommendation from machau people, SOP and yes, connections - all of these things. Having stellar internships also helps.
A: Personally, I think Masters have become very devalued. We live in a world where knowledge is freely flowing. There is no ‘knowledge’ that is specially stored in these degrees - it is the connections and the stamp of a good university people go to these colleges for. I didn’t think the ego boost is worth so much money. But if you have dispensible money and you want to do Masters for exploring new culture, pursue something very special, are evry sure that you want to work in a certain country or you know that the alumnus network is strong for you to go where you want in life, then go ahead. But doing masters from a fancy place just because society thinks its cool is a very very bad and an expensive idea.
P3: Hi! How to get into PhD in MIT. I’m 34 years old and been in academics all my life, no industrial experience
A: So I might not be the best person to answer this because I think of myself more of the startup guy than an academic guy, but I think the key to a good PhD is the right mentor who aligns with your goals and is willing to support you. Writing to professors about your interests and experience would be great. Attending conferences in US where you can network with Professors directly is a good option. MIT also has some great tie ups with Indian companies like TATA where they can sponsor PhD students working on topics crucial for developing nations. Reaching out to these institutions would be great.
P4: When you were a student, how did you takeout time to start and finish a side projects? It’s easy to start something but finishing it is the hard part for me. Or putting this another way, at what point do you think we should stop working on a side project because it’s taking too much time. I’ve only managed to finish only a few but it’s just demotivating when I look at the number of dead projects that I have :p
A: Make a goal towards which you are working. For me, the goal was getting in to MIT Media Lab, I would think about it while sleeping, while dreaming, while walking. It motivated me to do things. I knew that no matter how many ideas I have, it is only the finished ones that will go to my portfolio. At one time, I was just filling the grid boxes with thumbnails of my incomplete projects that would motivate me to finish them. Once I finished one after a lot of hard work, I was addicted to keep going. So build that momentum, align yourself to some external goal towards which you are working. In my initial years, it was just about impressing my friends with my ideas - find an external drive.
P4: When looking at it generally, how much of a difference do you think college(high ed) makes in a Student’s life? Do you think graduating from a not so good college(or should I say where the education is bad) pre 2015 vs graduating from that same college today in 2019 is different because we have so much of great learning materials on the Internet today that college should not matter at all?
A: I love this question. Yes, the world has changed sooooo much! It matters so little where you studied if you are in the startup world, maybe except for impressing few snobby VCs. YouTube, Coursera, all of those have shattered the walls academia had created. There is no reason to cite your college as an excuse. If you are smart, ambitious and hard working,(and are blessed to be born in a stable family in a stable country - which I hope you all are) this is the best time to be alive.
P5: I would like to ask that how to choose which particular field I should master in by keeping in mind about the scenario of the future market?
A: Three things - what you love doing, what pays you well and what you are good at. You may love aerospace and it even pays well but if you arent good at it, there is no point. Try to find things that hit all these boxes. Now, what pays you well is a subjective thing, there is no way of being sure but the world is going in a direction, there are some things like AI, creativity, data, service based industries, Electric vehicles and blockchains which are booming - see what connects with you in these emerging fields. Don’t stress out too much, as a thumb rule if you can code, you are guaranteed on the pay well aspect - the other two checkboxes are for you to infer.
P6: How much time we should devote to do projects or learning in one field may be it is data science or block chain? When to move to other field?
A: You cannot learn or even master a field while you are in college. People have been working on these for decades, you surely cant master these in a year. So I’d advise the best investment of your time in college should be to learn how to learn. Realize you will never master a thing fully but you can learn the basics so well that it helps you even when you are doing something unrelated. It is not the same as jack of all trades - it is knowing the basics of many fields very strongly so that you can use them if you have to.
P7: If you are doing your project in data science is it necessary to publish paper in reputed publications?
A: Ugh, I personally haven’t read more than 10 papers in my life. I find them very boring. It isn’t the medium that is important but your work. I honestly just write Medium articles, it has helped my work reach to more people. I find the academic paper wala language to be very terse and unnescessarily complex. It is one of the reasons I did not pursue academia ahead. If you are sure that academia is your thing, then yes, writing papers and getting them published is very important, but if you are doing a startup or are applying for jobs, its not important enough. Reading papers will definitely help you sharpen your mind, but not necessarily writing for publications.
P8: What are your views on opensource in AI? Can you give some inputs and trends about it! Also if you share any cool project you have worked on and YouTube channel link.
A: Hey Sourav! Open Source is responsible for probably 80% of the innovation happening today in the world. OpenAI, the initiative by Elon Musk is sharing so much of quality resources. Side note, all of you should be on Twitter and follow these guys - ElonMusk, Open AI, Andrej Karpathy, etc etc - you will get so much of cutting edge research work right in your feed. You can have a look at my projects at https://iashris.com - I think the Facebook network visualization (it has a video on YouTube) is probably the coolest one, followed by https://13x29.github.io
With that and some thank you notes, Ashris left the chat.
We thank both the participants of the AMA and Ashris for carrying out this extremely informative exchange!