SDS 505: From Data Science to Cinema

This is episode number 505 with Hadelin De Ponteves, machine learning educator to millions and Bollywood actor. 

Jon: 00:12

Welcome to the SuperDataScience Podcast. My name is Jon Krohn, a chief data scientist and bestselling author on Deep Learning. Each week, we bring you inspiring people and ideas to help you build a successful career in data science. Thanks for being here today, and now let’s make the complex simple.

Jon: 00:42

Welcome back to the SuperDataScience Podcast. We’ve got a fun and nostalgic one for you today with Hadelin De Ponteves. Hadelin is perhaps best known to listeners of this program as a renowned educator on machine learning. There is a long-standing partnership with the SuperDataScience company that bankrolls this very podcast. Hadelin has created and presented 30 extremely popular Udemy courses on machine learning topics like deep learning, machine vision, natural language processing and artificial intelligence. In all, his videos have sold more than two million copies, enabling Hadelin’s work to impact at least 1.4 million unique students so far. Prior to his epic creative period publishing machine learning courses, Hadelin studied math, engineering and AI at the University Paris-Saclay in Paris, France, a world-leading mathematics institution. And he also worked as a data engineer at Google. More recently Hadelin has written a book called AI Crash Course and was co-founder and CEO of BlueLife AI, an ML training and consulting company. 

Jon: 01:52

In today’s episode, however, we’ll primarily focus on Hadelin’s recent shift toward acting in mainstream films. Specifically, we’ll discuss the characteristics that enable an outstanding data scientist to excel in any pursuit. We’ll talk about how to cultivate your passion and achieve your dreams no matter what they might be. And we’ll talk about trends, similarities, and differences between the Eastern Bollywood cinema industry and the Western Hollywood industry. Early on in this episode, we do get a little technical discussing machine learning topics like the TensorFlow developers certificate and software modules for deploying deep learning models into production systems. But the episode opens up into a broadly inspiring and insightful conversation that should be enjoyable to anyone looking to find their passion and execute on it. This is a cool one, are you ready for it? Let’s go. 

Jon: 02:54

Hadelin, welcome to the SuperDataScience Podcast, it’s amazing that you’re here. I’ve known so much about you over the years, but we’ve never actually met. So we’re kind of meeting on air, there will be audience members out there who know you even better than I do. But I’m sure everyone is looking forward to getting caught up on what you’ve been up to lately. Hadelin how are you, where in the world are you? 

Hadelin: 03:19

Thank you so much. I’m very happy these days and I am in Dubai. And Jon, it’s so nice to meet you. Kirill told me so much about you. 

Jon: 03:30

Well, yeah. Truly it is an honor. That is mostly how I’ve heard about you is through the amazing things that the previous host of the show Kirill Eremenko has said about you. So you’ve had three SuperDataScience episodes. 

Hadelin: 03:44

Yeah. 

Jon: 03:44

The most recent one was number 327, which was in January 2020. It was on Data Science trends, looking forward over that year. And you and Kirill have collaborated on so much over the years, not just a few SuperDataScience episodes. But you guys work together on the SuperDataScience company, on BlueLife AI, which we can talk a little bit about. You’ve collaborated on dozens of best-selling Udemy courses on machine learning and AI, some of the best-selling Udemy courses of all time. So it’s funny you guys know each other so well, but yeah, here we are getting to meet each other. So Hadelin, what have you been up to since you were last on the episode? I understand there’s been quite a bit. 

Hadelin: 04:38

Yes. Okay. So yeah, indeed, beginning of 2020 was quite a while since the last podcast episode, so we have a lot to catch up. All right. So I’ve been up to two big projects. The first one is a related to data science and the other one is not related to data science. So let’s talk about the first one. The first one is a big TFC TensorFlow certification course, that I took a very long time to make because it’s basically a huge course, which we therefore released not on Udemy, but on our own website, which is a tfcertification.com. And it’s basically a course that prepares students to pass the TensorFlow certification exam. So basically what we do in the course is, we of course explain the whole theory of deep learning because that’s what you do with TensorFlow. 

Hadelin: 05:34

And we give a lot of practical activities on how to solve deep learning problems with TensorFlow. And of course we do that while covering all the different branches of the exam. Which includes, computer vision, natural language processing, time series prediction. And we even go beyond the scope of the exam because we also teach about TensorFlow light, which you can use on mobile, Raspberry Pi, whether it is iPhone or Android. And also we go beyond the scope by teaching TensorFlow with JavaScript. So it’s a huge course in 25 parts, which basically cover the whole TensorFlow ecosystem while focusing on the exam. And the plus of this course with respect to other courses, because you can find many courses on how to pass the TensorFlow certification exam, is that we give exam sample. 

Hadelin: 06:33

So we basically, literally took the exam several times. And we got familiar with all the exercises and the way to provide the solutions. And we basically offer some exam samples of the real exam, not the exact same questions of course, but some similar questions, and we give the solutions. We help them understand how to provide the solutions, code the solutions from start to finish. So that when our students take the exam, they’re completely familiar with it and they will basically get very similar questions. So they will, without a doubt solve the questions and succeed the exams. So that’s a big, big project we have been working on for quite a while. It’s done now, it’s out. I already did- 

Jon: 07:21

Can I ask you a couple of questions about that project before you move on to the big second one? 

Hadelin: 07:27

Of course, yeah. As much as [crosstalk 00:07:28] you want. 

Jon: 07:28

Yeah. So this TensorFlow certificate, I’ve known about it since the autumn, when it was launched and I’ve always thought it was a brilliant course. 

Hadelin: 07:35

Thank you. 

Jon: 07:36

So first of all… Yeah, not only because I agree that these subjects that you’re covering machine vision, natural language processing, time series analysis, these are the big application areas of deep learning and so somebody can get very much up to speed on the most important parts, but then also to include how you can deploy these TensorFlow models into lots of different situations. This is one of the most powerful things about TensorFlow and still today what makes it, for example, if you’re doing production implementations of TensorFlow models, of deep learning models, why it might be a much better choice than PyTorch for you because of these things, like the TensorFlow light library that you mentioned, for deploying into phones for Raspberry Pi. And very cool, TensorFlow.JS for deploying into people’s browsers, so that their own browser can run the model. Super cool. 

Hadelin: 08:35

Yes. And it’s funny that you mentioned PyTorch because to be honest, at the very beginning when all these platforms were out, there were basically TensorFlow and PyTorch, and I had a preference for PyTorch. I even say that to my earlier courses. But since TensorFlow 2.0, I’m a complete fan of TensorFlow, and now I prefer TensorFlow. 

Jon: 08:57

Oh, cool, man. Yeah, I can understand that. The TensorFlow-2 release was a big deal. It made it so that we had this imperative execution very much like PyTorch. I think the TensorFlow developers felt PyTorch was so easy to use, you could get results so quickly and intuitively in even like a Jupiter notebook, in a way that you couldn’t with TensorFlow 1, and so TensorFlow 2 was a big push to get aligned with that. Now, I understand [crosstalk 00:09:26], that TensorFlow certificate. I understand that a really big thing about it Hadelin is that, it allows people to get this… It’s a Google certified, certification, and it’s very difficult to get, there’s only a small number of people on the planet that have it, right? 

Hadelin: 09:45

Absolutely. But it’s not to be confused with the Google Developers certification, which is even harder to get, very, very few people have it. It’s like for example, I have a great friend in Serbia who we actually work together, and he’s the only one in Serbia to have it because it’s so difficult. In France, I think only 10 or 20 people have it, it’s basically the hardest IT certification exam to take, so not to be confused with it. But still you’re right that the TensorFlow certification is quite hard to get, that’s why we made a very exhaustive course on this and with a lot of practical activities and exam samples, for students to practice and be completely ready when they take the exam. But yeah, now quite many developers have it and they benefit a lot from it because basically the major benefits of having this certification is that you level up. You level up not only in your skills, but in your job. And by having the certification also, you get a higher chance to land a great IT job in a top firm. So that’s the major benefit. 

Jon: 10:57

Nice. So just help me be crystal clear on this, so that I’m understanding, that whether there’s two or three different certificates that we’re talking about here. So obviously there’s the TensorFlow certificate course that you created, and that is available from SuperDataScience at www.superdatascience.com. There’s the Google developers, kind of this TensorFlow certificate that you’re saying is very difficult to obtain, but your course is like a stepping stone to be able to obtain that very difficult certification, right? 

Hadelin: 11:29

So by very difficult, are you talking about the TensorFlow certification or the Google developer certification? 

Jon: 11:37

Yeah, I guess so… Yeah, this is what I’m trying to understand. So then there’s three that we’re talking about. There’s- 

Hadelin: 11:43

No, there are only two. 

Jon: 11:45

Okay. 

Hadelin: 11:45

There are only two certifications. 

Jon: 11:45

Okay. That’s it. 

Hadelin: 11:47

So basically there are only two certifications. There is the TensorFlow Developers certificate and the Google Developers certificate. And the one extremely hard to get, and you even have to take 15 interviews to get it, with Google partners that check in super depth, your knowledge, so extremely hard to get, and that’s the Google Developers certificate. So that has nothing to do with TensorFlow Developers certificate, which my course is about. Yeah. So the course that I have prepares students do a difficult exam, but not too difficult once you practice and all. So that’s why more people have it [crosstalk 00:12:25]. Yeah. 

Jon: 12:26

Okay. Now I understand. That’s perfect Hadelin, thank you for clarifying for me. 

Hadelin: 12:30

No worries. 

Jon: 12:32

And hopefully that’s useful for the audience too. 

Jon: 12:39

Eliminating unnecessary distractions is one of the central principles of my lifestyle, as such, I only subscribed to a handful of email newsletters. Those that provide a massive signal to noise ratio. One of the very few that meet my strict criteria is the Data Science Insider. If you weren’t aware of it already, the Data Science Insider is a 100% free newsletter that the SuperDataScience team creates and sends out every Friday. We pour over all of the news and identify the most important breakthroughs in the field of data science, machine learning and artificial intelligence. The top five, simply five news items. The top five items are handpicked. The items that we’re confident will be most relevant to your personal and professional growth. Each of the five articles is summarized into a standardized, easy to read format, and then packed gently into a single email. 

Jon: 13:35

This means that you don’t have to go and read the whole article, you can read our summary and be up to speed on the latest and greatest data innovations in no time at all. That said, if any items do particularly tickle your fancy, then you can click through and read the full article. This is what I do. I skimmed the Data Science Insider newsletter every week. Those items that are relevant to me, I read the summary in full. And if that signals to me that I should be digging into the full original piece, for example to pour over figures, equations, code, or experimental methodology, I click through and dig deep. So if you’d like to get the best signal to noise ratio out there in data science, machine learning and AI news, subscribed to the Data Science Insider, which is completely free and no strings attached at www.superdatascience.com/DSI. That’s www.superdatascience.com/DSI. And now, let’s return to our amazing episode. 

Jon: 14:35

Okay. So that was your one really big project since the last time you were on the episode. And I know you wanted to talk about another really big project, but I think I know what that is. And so just before you get to that one, I want to also mention that there’s some really big things that you’ve done in the data science space. You’ve got this AI Crash Course book that was published by Packt. 

Hadelin: 14:57

Yes. 

Jon: 15:00

And yeah, you have in general, not just this TensorFlow certificate, but dozens of courses on AI and ML that are available in Udemy. And I think there’s probably been some development on some of those since we last spoke as well. So I just wanted to mention those so the audience is aware of- 

Hadelin: 15:18

Yeah, thank you. 

Jon: 15:18

Yeah. 

Hadelin: 15:19

Yeah, no, thank you for mentioning those. Yeah, for sure. Actually, I released my book at the end of 2019, December 2019. So indeed I think I announced it on the last episode, but it’s always good to remind it because it’s a great complimentary resource to the courses. And of course we do TensorFlow inside, we do a lot of practical activities and that’s very useful for the students. I’m actually getting some very nice feedback, so I’m very happy about that. Yeah. So can I move on to the second big announcement? [crosstalk 00:15:56]. 

Jon: 15:57

100%, Hadelin. 

Hadelin: 15:57

Okay. 

Jon: 15:58

I know you’ve been waiting and I keep delaying it, so let’s get to it. I’m excited. 

Hadelin: 16:02

Oh yeah, let’s get into it. Because there’s actually a big parallel between the first… Well, what I’ve been doing for the past couple of years, transferring knowledge and this second announcement. So basically I’ll say it right away and then I’ll develop. The second announcement is that I am doing cinema. I was trying to look for a cool way to say it, but I am doing cinema. I am acting in films, in India, in Bollywood. 

Jon: 16:35

Oh, yes, it is Bollywood. I knew it was India, and I was like, “I suspect it’s Bollywood,” but I don’t know enough about it to be like, “Is it definitely Bollywood?” [crosstalk 00:16:43] so it’s Bollywood. I’m even looking forward to hearing about what that culture is like, but yeah, I’ll let you talk. Tell us about it. 

Hadelin: 16:49

Yeah. So what I really want to say first to highlight this parallel, is that there is a big common denominator between making courses and making films. And that comes from my inspiration, it’s kind of the same inspiration. The inspiration is to share something, right? So with my courses, what I’ve been sharing is knowledge. I had this huge inspiration to share my knowledge in IT, AI, machine learning and all. And now by making films, by acting in films, I’m also sharing something, which is not knowledge, but this time emotion. So I have this huge inspiration to share emotions. And so basically I’m switching from sharing knowledge in my courses to sharing emotions in films. And I think that comes from the same deep inspiration that I have in me. And maybe at some point I will also share knowledge in the future movies hopefully. Because movies are of course about conveying emotions from your character to the audience, but also you can spread a strong message in your films and you can transfer knowledge through that message. 

Hadelin: 18:12

And of course, then if you’re doing some documentary movies or historical movies, you’re sharing even more knowledge. So I think this is not a coincidence or this is meant to be that I’m switching from a data science to films. By the way, I’m not completely switching from data science to films, I’m just doing that on the side. And I’m still very inspired to share knowledge, it’s just that I shared all the possible knowledge I could, and I’ve shared basically all the skills I had in AI, but of course I’m still staying in the game. And as soon as there is some big, new updates in AI or big progression or a bigger new topics to teach within the AI ecosystem, I’ll be very happy to make another course and transfer a new knowledge. 

Jon: 19:08

Very cool. 

Hadelin: 19:08

So right now, the top of my inspiration is to share emotions. And so that’s why I embarked myself upon this Bollywood film. And this was an amazing experience, we have just finished the shooting. Now the movie is in post-production and it will be released in October hopefully, that’s the best ideal scenario, October. But there’s always some delays, so I suspect it could be released by November. But wow, I had so much fun doing this. And when I was acting I really had this sensation that I was… Because my character actually has a lot of strong emotions in the movie. So it really resonates well with my why of doing films, of my inspiration. So yeah, that’s the second announcement, I hope the SuperDataScience community will be excited to see the movie. For sure, it will look very weird for them because they have been used to listening to my voice only in the courses. And suddenly they will see my face and see me doing some dramatical behavior on film, so this will be very weird for them. But I hope they will enjoy the movie. 

Jon: 20:29

Nice. Well, so Hadelin, can you tell us the title of the film or anything about the plot? 

Hadelin: 20:34

Yes, of course. So the title of the film is a Faraaz, it’s spelled F-A-R-A-A-Z. And it’s directed by one of the top Bollywood director’s, who’s named Hansel Mehta. He’s a great director. I really enjoyed working with him because, it’s like we had a common vision on my scenes. He had a vision and I was able to deliver what he wanted. So but of course that that is thanks to his great guidance, and greater communication with the actor. So I’m really grateful and feeling totally blessed to have worked with him, and I’m so excited about the movie now. It’s going to be released soon. 

Jon: 21:28

Nice. So Hadelin, this makes perfect sense to me, now that you’ve explained everything about what you’ve been doing over the last two years. When I first found out that you were working on cinema, I was like, “What’s the connection? How did this happen?” But now that you’ve explained it, it makes perfect sense. And so you are an outstanding presenter and you think of ways of, “How can I best convey something to the audience?” And that’s what has allowed you over these dozens of courses in the last few years to impact many, many millions of people around the world, and why your courses are so beloved. And so it doesn’t surprise me that you’re able to transfer that capacity to conveying emotions, and maybe in the future, other kinds of messages through film. And it also makes perfect sense to me, given how incredibly productive you were over your years working in data science. I can’t even imagine how many hours every week were spent creating content because it’s just prolific. 

Hadelin: 22:36

Oh yeah, so much work. I remember when I made my first courses, I was literally spending all my time making the courses. I did a complete sacrifice on social life, actually. I was even working 24 hours on the weekend. And actually I’m quite well known in this business for having had this three hours sleep routine. 

Jon: 23:02

Really? 

Hadelin: 23:03

If you listen to the first SDS podcast episode that I did, this was the main topic of the episode. Kirill highlighted that I only slept three hours every night or even during the day, because each time my sleep was shifted by one more hour the next day. Because I was basically going to sleep when I was completely exhausted, like with 0% energy, I slept three hours, and then I went back to my office to make my courses. So yes, it was incredible hard work. And I’m glad you mentioned this because this is another common points with the movies. Making this movie was extremely hard work as well. We woke up at 5:00 AM, we got on set at 6:00 AM and we shot until 11:00 PM. 

Hadelin: 23:52

So it’s like the same incredible, hard work process that I went through. And so, there is definitely another parallel with that. And I can even highlight another big parallel that you will enjoy to hear. It’s the fact that in my courses, when I shared my knowledge, I was already very theatrical. Actually, one of the reasons why my courses were appreciated is because I was very engaging while explaining the AI topics. I was very excited, always with some big voice modulations, which is a technique we have to do in films, for example, when acting. So I was already very theatrical, and so that definitely helped as you said, presenting something in the movies as well. 

Jon: 24:44

Nice. That makes perfect sense to me. I didn’t know about this three hours sleep routine. I’m guessing that if it was your first episode in which you were talking about that, I went and I looked it up in the past episodes, it looks like that was episode number two of the SuperDataScience show- 

Hadelin: 25:01

Yes. 

Jon: 25:02

… right at the beginning, in 2016. 

Hadelin: 25:05

Absolutely. 

Jon: 25:06

So people can scroll all the way back to get to that episode number two, crazy. That’s more than 500 episodes ago. 

Hadelin: 25:19

Wow. Great work also on the podcast. 

Jon: 25:23

Yeah. 80 or 90% of those are Kirill. And if I had been left to create a podcast, I don’t think I would say we should do two episodes a week, on top of already having, in Kirill’s case or my case, other jobs, where this is something we just do on the side. Twice a week seems crazy. But with Kirill having done that, I felt there was this bar that I had to meet. And it’s very interesting, I’m sure this is something similar to what you’ve experienced with all of the accomplishments you’ve had, is when somebody presents you with this challenge and you have this example of somebody who has done it, you think, “Okay, somehow I’m going to make this work.” And it’s turned out. I’ve been able to make two episodes a week since I took over in January, which if [crosstalk 00:26:20] I didn’t look to my own devices, I never would have done. 

Hadelin: 26:24

Yeah, that’s amazing. Yeah. And it’s a great podcast, very successful. I actually still get more or less every day, LinkedIn connections from people who listen to the podcast episodes and who I mentioned the sleep again. I get [crosstalk 00:26:42]. Basically, it’s always the same thing. They ask me if I still sleep three hours per night, which I don’t anymore. 

Jon: 26:51

Okay, yeah. 

Hadelin: 26:51

It was only during the launch. 

Jon: 26:53

That’s what I wanted to know too.

 

Hadelin: 26:55

Yeah, no. It only lasts one year actually, because then my body could not follow. And plus it has to do with passion, I was extremely passionate to make my first courses. Of course, because you’re launching your business, you are on your purpose, you create something. So you’re extremely passionate and that’s why I could get up my bed after only three hours of sleep. 

Jon: 27:22

Wow. I’ve read a lot about these kinds of techniques. So my academic background is in neuroscience and part of that is studying how we sleep. 

Hadelin: 27:34

Very interesting. 

Jon: 27:35

And I remember reading at that time that it is possible to train people, to do this, to be… So about three hours, that’s the minimum before really bad things start to happen physiologically, less than three hours. And so I don’t know if you know this from the time that you were doing it, but a lot of scientists think that the reason why human sleep for eight is because that historically when we were hunter, gatherers, tens of thousands of years ago, when our species evolved, we needed about 16 hours a day to collect enough energy resources to survive. And then you spend the rest of the eight hours asleep to protect yourself, to avoid getting in trouble. And the reason why this hypothesis is so strong is that when we look at other animals, basically the amount of time that they sleep corresponds to the time that they don’t need to be gathering food. So if you are a horse or a cow, you’re a big animal, but you’re basically only eating plants. 

Hadelin: 28:45

Which I am by the way. 

Jon: 28:48

That’s funny, it’s the cow sweater. So that makes perfect sense, that’s why you need to sleep so little. So they need to spend- 

Hadelin: 28:56

That’s so interesting. 

Jon: 28:57

Yeah, they need to be eating grass all day. 17, 18, 20 hours, maybe of just consuming grass, and so they can’t sleep for eight hours, 10 hours. But a cat or a dog, they can get all of their nutritional requirements in just a few hours a day, and so then they sleep the rest of day. 

Hadelin: 29:16

Oh, okay. That makes sense now. That’s very interesting, I did not know that. 

Jon: 29:22

So these scientists think that mammals need at minimum about three hours of sleep, in terms of your physiological processes, in terms of their base requirements, but all of the rest of that sleep is dependent just on your energy gathering needs when your species evolved. 

Hadelin: 29:39

Oh, okay. 

Jon: 29:40

So from that, these scientific studies were saying, “Okay, well then how do we get humans down to three hours?” And one of the ideas, maybe this is similar to what you did, I’d love to know, is reducing the amount of sleep by 15 minutes a night, every week to get you from eight hours, and then after four weeks you’re down to seven hours, and then after another four weeks you’re down to six, and gradually you can get down to three. But so that’s one approach that I’ve heard works pretty well. But no matter what, and I’d love your feedback on this as the first person that I’ve ever spoken to that’s only slept for three hours a night. My understanding from those studies, of these people who would get down to three hours of sleep every night, their cognitive performance and their physical performance was as good as when they were sleeping their full amount. 

Hadelin: 30:29

Oh yes, for sure. 

Jon: 30:30

But they reported feeling very tired all the time. 

Hadelin: 30:35

Oh really? 

Jon: 30:36

That was the main downside, yeah. You didn’t have that? 

Hadelin: 30:39

No. But I forgot to mention something. So I was basically sleeping my three hours per night or per afternoon or per morning, but I was also doing some power naps. So indeed, at some points when I’m feeling a huge heavy sleep need. And then when it happens, I would just do a 10 minute power nap, but not more than 10 minutes, and I would get back on my office. And I was having maybe five power naps per day, plus besides the three hours sleep. 

Jon: 31:18

I see. 

Hadelin: 31:18

So I guess that was my way of surviving. But again, I think it has to do with passion. Because I was extremely passionate, I think the passion was stronger than the need of sleep. And therefore, I don’t remember myself being very exhausted all the time or very tired. 

Jon: 31:35

Oh, nice. 

Hadelin: 31:35

Yeah. [crosstalk 00:31:36]. 

Jon: 31:37

And maybe the big difference there is that, in your case, this sleeping so little was driven by an internal desire. Whereas in any of these studies, it is something that’s mandated by the study. It’s like, “Okay, next week, you’re going to reduce your sleep another 15 minutes.” And they’re like, “Oh my goodness, another 15 minutes, I’m already so tired.” Because they don’t necessarily have the passion backing it up. 

Hadelin: 32:01

Yeah, exactly. I really believe it has to do with this also. 

Jon: 32:06

Nice. So, okay. I understand very well now these parallels between the content creation that you’ve been doing over the last few years and that it sounds like is inevitable that it’s going to happen again- 

Hadelin: 32:19

Yes. 

Jon: 32:19

… as well as what you’re doing in cinema. And so, it’s this idea of passion driving what you’d like to do. And so how did this route present itself to you? How did you decide, or how did you become confronted with this opportunity to create film? And how, or why specifically did you choose Bollywood is as the place to have that passion expand? 

Hadelin: 32:52

Okay. So thank you so much for asking this question, I love this question because I’m fascinated by the answer. So I’m a very rational person but I also believe in things you can’t yet explain. And so the first answer to your question is, how I got into this is, is by, I think the law of attraction, which you can’t really rationalize really well today, but which I really believe. It’s the fact that when you really think about something it happens, basically you become the product of your thoughts. And I had been thinking quite for a while, for a couple of years to do this, to do films. So once again, it comes from my deep inspiration. At first my inspiration brought me to share knowledge, but later it brought me to sharing emotions. 

Hadelin: 33:48

So I had been thinking more and more about this, and there’s actually an anecdote that Kirill will love to hear. We basically went on a motorcycle trip in New Zealand and Kirill at some point kind of woke up that inspiration, because he said, we were filming this very cool motorbike scene in the mountain on off-road and he said, “Wow, this looks like a movie, you could become a great actor.” And kind of woke me up, woke up that inspiration. And so since that moment… by the way, thank you, Kirill, because it could be thanks to him all this. Since that moment, I started to think more and more about this. And so more and more of my thoughts were about cinema and with the law of attraction, which is that you become the product of your thoughts. By basically thinking about this, visualizing it and asking for it, well by the mystery paths in life, I ended up meeting someone at some point that is connected to the Bollywood industry. 

Hadelin: 34:59

And by one mean to another, I got to audition for this movie and got the part, which was amazing. So I said, “Okay, well, this is a great opportunity.” And I ended up making this film, so I was super, super, super grateful. And now that I did it and that I completely loved it, I want to do more films. And now this is going to lead me to your second question, why Bollywood? So it’s just because my opportunity led me to Bollywood. But in fact, I would be super happy to do also Hollywood or even British cinema or even French cinema. So it’s just the randomness of life that got me to Bollywood, and I really enjoyed that experience. So I would love to do some more Bollywood movies, but I’m completely open to other film industries. And also, there is a rational way to explain how I got into Bollywood. It’s not only the law of attraction, it’s also the fact that I live in Dubai and that’s only three hours away from Bollywood, right? 

Jon: 36:25

Right. 

Hadelin: 36:25

So that’s the thing. I know a lot of Indian people in Dubai. My girlfriend is actually Indian and she’s also an actress, she’s a stage actress switching to films. But all these elements led me to Bollywood. And that’s definitely partly due to the fact that it’s so close to Bollywood. For example, if I was in Mexico, maybe I would have ended up in Hollywood. We don’t know, but yeah, that’s the thing. There’s both a rational explanation to this and an irrational or a spiritual explanation to this. I don’t know how to say it best, but yeah, that’s basically how I think I got into this. 

Jon: 37:11

Nice. That is super cool to hear that, Hadelin. And so just so that I understand a little bit better about Bollywood. So like Hollywood, it’s a physical place, right? I guess? 

Hadelin: 37:23

Yeah. 

Jon: 37:23

But the name is kind of inspired by Hollywood, right? 

Hadelin: 37:26

Yeah. 

Jon: 37:26

And so what are the differences between Bollywood cinema and Western cinema? 

Hadelin: 37:32

Okay, yeah. Thank you. Great question again, because I actually asked it myself. And I had to ask myself this question in order not to be a fool in the Bollywood movies. So basically at first to answer your question, yeah Bollywood is basically Bombay. It’s the city of Bombay, but because the city of Bombay is full of film studios, we also call it Bollywood. Okay. [crosstalk 00:37:57] they even call it the industry over there, they call it the industry. Because this is the major film industry, they don’t call it film industry, they just call it industry. So yeah, that’s the thing. Now, the big difference between Bollywood and Hollywood is that in Bollywood… how should I say this? They are much more expressive with their way of acting. 

Hadelin: 38:25

It’s much more theatrical. The expressions are much stronger, a little bit exaggerated than Hollywood, where the expressions are maybe more natural, more subtle and less theatrical. So that’s the major difference I understood. However, I’m seeing a smooth transition into the Western style. Because I watched a lot of Bollywood movies, and so indeed in a lot of them, especially the old ones, or the ones from 10 or 20 years ago, I see in these some very exaggerated or theatrical or strong expressions in movies. But in the latest ones, for example, Kapoor & Sons, or Barfi, which are Bollywood movies, well, I could see some much bigger similarities with the Western films, with more subtle expressions, natural expressions. 

Hadelin: 39:30

And so I think I’m happy about that because the director told me that I was a more subtle actor than a theatrical actor. So it’s quite perfect for me if there is this transition. But anyway, as we said, I’m open to every industry, so hopefully I will be in other industries as well, not only Bollywood. But yes, that’s the major difference I noticed. And of course the other big difference, obvious difference is that Bollywood movies are in Hindi and Hollywood movies are in English. [crosstalk 00:40:02] We actually see more and more of English lines or scripts in Bollywood movies, that’s another thing. So I think there is kind of a westernization of Bollywood movies right now. And that makes sense, because right now the world is becoming more and more connected. Indian people are now all on Netflix, so they have access to the Hollywood movies, and so they start to like the Hollywood movies more and more, and therefore the Bollywood movies are getting closer to them. So yeah, I’m really observing this transition and it’s fascinating. 

Jon: 40:41

Do you speak Hindi, Hadelin? 

Hadelin: 40:43

Yes, I do. I’m learning- 

Jon: 40:45

Okay. 

Hadelin: 40:45

… Hindi and I speak the basics, beginner level. Yeah. But- 

Jon: 40:49

So in this film Faraaz, were you speaking English or Hindi? 

Hadelin: 40:52

No, all my lines are in English. And I even say a line in French. The director asked me to do a line in French to show a certain kind of emotion. I don’t want to… how do we say? Spoiler? I don’t want to spoil it but yeah, I say a line in French. Yeah. 

Jon: 41:13

Nice. Okay, that’s cool. Well, everything that you’re talking about, this idea of law of attraction and that kind of thing, I think is very important and applicable to anyone in any career. So whether you’re listening as a data scientist or anything, this idea of your mental state changing, you have this idea because Kirill tells you in the mountains of New Zealand, that you could be in a film. And then you think about this more and more, and you can look for opportunities. And so an opportunity that would have presented itself either way, you don’t even notice it if you’re not looking for it, but if you’re thinking about it and you’re looking for it, and all of a sudden you accumulate these opportunities. A book recommendation that I highly recommend for the audience is The Artist’s Way by Julia Cameron, which is full of exercises that allow you to rekindle your creative self that you might’ve had as a child and that we tend to lose as adults. 

Jon: 42:16

And that book was huge in allowing me to transition from just a data scientist, typing code, to a data scientist, having more creative ideas, working with code. And then eventually that led to this creativity of creating lots of different courses and talks, and eventually this podcast, the SuperDataScience Podcast. So Hadelin, while working from a hotel lobby to get internet in Dubai, his battery ran out just as we were about to finish wrapping up the episode, but I got from him that his book recommendation other than of course his own AI Crash Course for you, data scientists, is The Alchemist by Paulo Coelho. So these are great books for learning about the law of attraction in your own life. 

Jon: 43:15

The Alchemist for example, is a fictional work, but it lays out these ideas of how you can yeah, follow your passion and find your purpose in life, so great recommendations. And then Hadelin, let me know that the best ways to stay up to date on what he’s doing, whether it’s in Bollywood or on Udemy or www.superdatascience.com is via LinkedIn. So he has a huge LinkedIn following there of tens of thousands of people and he’s also on Instagram. And yes, of course we’ll have links to both of those in the show notes for the episode. 

Jon: 43:59

What an inspiring character Hadelin is. I’m so happy to have finally had the chance to meet him. In today’s episode, we discussed the prestigious TensorFlow certificates available to data scientists and software engineers who would like to demonstrate their chops in deep learning. We talked about the parallel between making courses, where one shares knowledge and making films, where one shares emotions. We talked about how the same traits that make an outstanding data scientists like passion and hard work, can enable someone to succeed in any field of their choice. And we covered books to stimulate your own creative flame, or enable you to discover new obsessions and fulfill them with fervor. As always, you can get all the show notes, including the transcript for this episode, the video recording, any materials mentioned on the show, the URLs for Hadelin’s LinkedIn and Instagram profiles, as well as my own social media profiles, at www.superdatascience.com/505. That’s www.superdatascience.com/505. 

Jon: 44:58

If you enjoyed this episode, I’d of course greatly appreciate it if you left a review on your favorite podcasting app or on the SuperDataScience YouTube channel, where we have the video version of the episode. To support the SuperDataScience company that kindly funds the management editing and production of this podcast, without any annoying third-party ads, you could create a free login to their learning platform, at www.superdatascience.com, or consider buying a usually pretty darn cheap Udemy course, published by Ligency, an affiliate of SuperDataScience, that just did 800,000 copies selling Machine Learning A to Z by Hadelin himself, along with former SDS podcast hosts, Kirill Eremenko, or my own Mathematical Foundations of Machine Learning course, which is also available in Udemy. 

Jon: 45:46

Alright, thanks to you Ivana, Jaime, Mario and JP on the SuperDataScience team for managing and producing another amazing episode for us today. Keep on rocking it out there folks, and I’m looking forward to enjoying another round of the SuperDataScience Podcast with you very soon.