#89 Hadelin de Ponteves is a data scientist and an entrepreneur
⚠ The following transcript was automatically generated.
❤ Help us out, Submit a pull-request to correct potential mistakes
Hadelin de Ponteves 0:00
The most important thing that I needed during my entrepreneurship life was inspiration. I had it most of the time, but I lost it at some point and I can tell you when you lose it, you know things go wrong. You You don't believe anymore you don't. You're starting to doubt. And the only thing that will keep you inspired is to stay focused on your purpose to stay focused on your goal, you know.
Tim Bourguignon 0:34
Hello, and welcome to developer's journey. The podcast shining a light on developers lives from all over the world. My name is Tim Bourguignon, and today, I receive Hadelin de Ponteves. Hadlin is the co founder and CEO of blue life AI, which leverages AI for optimizing processes, maximizing efficiency and increasing profitability. Atlanta is also an online entrepreneur who has created educational e courses about machine learning deep learning, Ai, and blockchain, which have reached over half a million customers worldwide. Hadelin, welcome to dev journey.
Hadelin de Ponteves 1:14
Hi, thank you very much for having me here. I'm very excited.
Tim Bourguignon 1:18
It's my pleasure to receive you today. So let's go back to maybe your childhood and know you're going to tell me when did you develop this passion for science and technology?
Hadelin de Ponteves 1:28
Oh, very, very early actually. Because Beckett's cool Well, my my strongest subjects were mathematics and physics. So naturally, I develop some kind of a high interest for new technologies science. And, and it was until high school that I was sure that I really wanted to start a career in new technologies. So yeah, I guess it all started with my bad Mathematics
Tim Bourguignon 2:01
is something that runs in your family.
Hadelin de Ponteves 2:03
Yes, my dad is a scientist, it himself. He was an engineer. He was also very good at mathematics. Then the, he had a career in optics. So he built telescopes basically. So it was kind of a very specific branch of physics that he chose to, to go for. And mine is a very different mine is it? I'm passionate about it artificial intelligence, data science, machine learning, and the future in general. And even the new technologies like blockchain that fascinates me, even if it doesn't really have to do with artificial artificial intelligence. But yeah, we have this common passion for the root of science, which is mathematics
Tim Bourguignon 2:50
makes sense? It's funny how family lineage, kind of influence what you decide afterwards. Even if you try to cut from it. Sometimes catch up with you.
Hadelin de Ponteves 3:00
Yes, however, I seem to be the, the only entrepreneur of defending my parents had the classic, you know, careers like, you know, working for companies. And I chose very early to start my own business, but I guess it's also a matter of generation. I have a feeling and I, I observed that this new generation is more entrepreneurial.
Tim Bourguignon 3:25
This entrepreneurial mindset is something that you had already early on.
Hadelin de Ponteves 3:29
Oh, yes, definitely. Because very early, I started to do things differently. For example, I never felt in the mood of society when I was at school, you know, I never felt comfortable at school. I felt that this wasn't the right way for me to to learn and grow. So yeah, I actually always hated school. And and very early, I started to learn different topics and subjects on my own resources. So instead of following the program, that's cool, I was following my own programs that my, my home and and I was going to school because it was compulsory but still I was more comfortable learning on my own resources. You know, I was buying books, in bookstores, instead of studying on the books at school, I was watching a lot of online programs. So yes, I was really making my own path very early at school. And for me, that's kind of the early signs of becoming an entrepreneur because you know, being an entrepreneur is kind of doing things differently as, as the way people are conditioned by society, you know, to have a job and then you know, start a family and pay him have mortgage, buy a house, I wasn't going to all this so I want to live my own different adventure and yes, I ended up doing things very differently even after all, after after school,
Tim Bourguignon 5:05
I went to st question. Okay. And since since we're both French I kind of looked at your, at your studies in a different way. And so the engineering school that you did actually one of the, the most standard way of becoming an high high end engineer, so it doesn't completely match up with this with this idea of intrapreneurship. So, why did you choose to, to go this way? And then bifurcate?
Hadelin de Ponteves 5:34
No, actually, I, I ended up kind of randomly in this in engineering school because of my dream, to, to go for an American University. So I wanted to study in an American University in the US. So that's why I actually started study at a French University. Then I apply to those American universities, but unfortunately I got rejected from all of them. And so Okay, so I said, Okay, I worked hard for these universities. So I'm going to try the French engineering schools, which I did. And, and yeah, I ended up there without having it prepared. You know, from the beginning. Basically, after high school, my goal was to end up in an American University. And I couldn't do that. So, as a backup plan, I went to a French engineering school. Because, okay, I didn't feel it in the mold of the society. I didn't feel well, it's cool. But yet I was a good student. I had good grades. So yeah, I had the opportunity to, to land in a great engineering school. And of course, this is this was very beneficial to me. And this really helped. So after after that, I was lucky to land a good job at Google, which only last One year, after which I became an intrapreneur. But I knew all along that I was going to become an entrepreneur just that I didn't want to refuse the high quality of a French engineering school, plus a one year experience in a top tech company, it makes
Tim Bourguignon 7:16
things more glaring that then then then let's take a little bit longer, but on the on the intrapreneurship. And then we'll go on to take a little bit. How did this did this engineering school prepare you for this intrapreneurship?
Hadelin de Ponteves 7:29
Well, I wouldn't say it prepared me for my for this entrepreneurship because entrepreneurship is more something I had me all along. So I knew I was going to be an entrepreneur. What it prepared to me, for me was, you know, the content, the product that I created afterwards, because, indeed, during my engineering school, I studied all these advanced subjects and advanced topics, on on science on Artificial Intelligence on mathematics. And therefore, since my first business as an entrepreneur was to create online courses, well, you know, the content of these online courses was very high quality, thanks to all the subjects that I had studied in mind during school. So I guess it helps more in the in the creation rather than the mindset. Although the French and German school that I did is very entrepreneurial, you know, there is an entrepreneurship option actually, the the final year, but Funny enough, I didn't choose that option. I chose a Strategy and Finance option. So But yeah, I guess yeah, the mindset, I already had it. However, it really helped to create a high quality product. And I'm sure that it also helped me in some other ways, but it's hard to tell you know, Because you swallow so much knowledge so much, so much things from your four years at dangering School that at some point you don't know what, what what was useful in the end, but I'm sure that you know some some of the rules or some of the paradigms that I was that my mind was trained about were helpful then for my entrepreneurship journey,
Tim Bourguignon 9:27
when did you discover the world of data? If I can call it this way?
Hadelin de Ponteves 9:34
I would say I discovered it at my final year of my engineering school, the importance of it, because previously, actually previously, I I didn't plan to become a data scientist or to do data science or artificial intelligence. I actually plan to do a financial engineering. But then the hype for data science and artificial intelligence started to appear and I got cut into it. And I became fascinated by it. And so, yeah, that happened actually in the final year of my engineering school. And I remember that, at the beginning of the year, I still plan to follow a financial engineering program. But I changed at the very last second, meaning I went to, you know, the director's office to say, Hey, I have a feeling that there is something better for me than financial engineering. And that's something better was data science. So at the very last second, I changed my, you know, final year program to be data science. And I did well because otherwise today, I would probably account or trader and back and not really an entrepreneur, because it's hard to become an entrepreneur in finance
Tim Bourguignon 10:49
today, I guess. Do you remember what what tilted the scale or flip the switch?
Hadelin de Ponteves 10:55
Yes, it was a I went back to school on the final year. And so I started to, to network and to talk about things with everyone. And some people told me about this new this new thing, this new data science program and this new because actually the data science program was new at the time, I actually did the first year of the data science program when it was created in my school. So there was this new program and at the same time, the news talked more and more about artificial intelligence and data. So yes, I had a feeling that there was an opportunity that was right for me to see how did he did you start learning about this? Did you just attend the program and learn through the program? Or did you do some some learning on your own to to procreate yourself this this topic? No, no, I never changed. I never changed. Of course, even if I actually, you know, change the program on purpose to you know, to follow the classes of the new program. Well, yes, there was still the old me and my old self that to one of the best books and, and to study on my own on my own resources. So yes, I did this I bought the best. I bought the best books on artificial intelligence, machine learning data science. And I mostly studied on these books rather than on the school programs. Did you remember some of some of these books the one that they were really forming for your, for your learning? Oh, yes, of course. So I read basically what I what I did was pretty simple. I went I went on Amazon and I, I typed on machine learning or artificial intelligence. And I took the most well the highest reviewed ones. And and I read them so yes, I read of course a Python machine learning, machine learning and our artificial data science from scratch by Joel gross and artificial intelligence. Now, deep learning Well, the classic deep learning book, reinforcement learning an introduction by Richard sudden, well, all of them all, well, I probably read like 22 to 30 books on artificial intelligence and machine learning, which was really helpful for me then to create online courses. Because what I did was, of course, to try to differentiate from all these books or to, to make something more complete than then. And that's how I ended up to to create some kind of the Bible of machine learning. And yeah, the disc came from all these are resources that I read. And this allowed me to differentiate quite well.
Tim Bourguignon 13:41
It took about this ecourses for a four minute yet, where did the idea come from,
Hadelin de Ponteves 13:47
um, after I stopped a Google, I actually had several ideas as business to start. And one of them was a online education. I actually wanted to create my own The online education platform, and I actually plan to create my own courses on the platform. So how did I have this idea? Well, I actually brainstormed with myself and realized that one of my core skills was a self learning, no self education I was really good at at self education, meaning, I was very good autodidact, I was very efficient at learning new things by myself, without the need of, you know, being guided or without the need of an institution like was cool. No, I could read and learn about tons of new subjects in a very efficient in a very efficient way. And all that by myself without any guidance. So that was my core skill, and I had identified it at the time. And so I thought that, therefore, relevant idea of a business for me would be to create online courses and create You know, an online education platform where I would provide my own creations, my own my own courses. And so that was my first idea. But then I found out about Udemy, I found out that this was a very successful online education platform that was growing massively, that was expanding massively across the globe. And therefore, my next idea was to not create my own platform, but just create my own courses and then put them on this existing successful platform. And that's how it all started. I created this first course with my business partner Kirill, Eremenko. And I got in touch with him through like an account manager Udemy and yes, I that was my first creation and, and it was an in when very successful. But yeah, to answer your question, the way I got into This was by identifying my core skill, which is a
Tim Bourguignon 16:04
very efficient self learn. Did you do something special to identify this core skill? Do you have a special retrospective that you do is yourself or brainstorming session that you do regularly or something like this?
Hadelin de Ponteves 16:16
Yes. Well, first I talked to that to people and advisors. So, you know, in France, there are these coaches that try to orientate yourself, like, guide you into a good path for your career. And so with one of them, yes, I actually, actually, you know, we actually assessed my, my skills and my, and my weaknesses, let's say, and so, yes, with all this conversation, well, one day I had one conversation with myself and, and I actually also read this book about By Dhingra juicy and in the book it's it explains how important it is to identify your course not only your skills, which I had that identify with these advisors, but also your core skill your best skill. There's also this great book called The big leap and and in the in the book explains that we all have a zone of genius, which is like the core skill, the best skill that you have. And so I that that kind of triggered me to to figure out what is really my zone of genius, what is really my core skill. And, actually, yes, that's how I really identify that my core skill was a self learning being a very good autodidact and with the ability to to learn new subjects very fast
Tim Bourguignon 17:48
on my own. That's really really strong equivalent to having realized as an inferior that early on, that's really cool. I took up too fast. To the E courses, actually, we we jumped over your, your job at Google. So how did you go from data science at engineering school or university level to being hired hired at Google and doing the work of a data scientist for one year?
Hadelin de Ponteves 18:17
Oh, well, that was easy. Actually. My entering school has had some partnerships with Google. So in the middle of the year, you get access to a list of opportunities of job opportunities with companies, partners of yours of the school. And so one of them with Google was Google. So I applied because the job was really interesting. And plus, it was Google. So I figured, you know, it was it would be a great first experience. And by the way, yes, I forgot to precise that it was compulsory for me to do this first work experience. So it's not like I only wanted to have a work experience before becoming an entrepreneur, but it was compulsory to do at least six months. experience at the end of the engineering program. So anyway, is there was this list of job opportunities? And and so yes, I there was Google among them. So I applied to Google one. I applied to also some other other opportunities in New York. And, and yeah, it turns out in the end, I was accepted to one of the opportunities in New York and the one in Google. So I chose the one in Google because this was the most closely related to data science. And this was actually data engineering to be more precise. And yes, I ended up in Google like that. So yeah, the reason why I say was easy, it's not you know, was not the the classic process of me applying to Google coming from nowhere and having to pass all the different steps of the interview. The interviews, no, this time you know, they knew from which program I was coming from, then you it was a very high quality program. On. So they had me come to Google for three, three or four interviews. And that was it. And the interviews were not that hard. So it was really from the help of the school network. And the way it was recognized as a high quality program,
Tim Bourguignon 20:17
you just made a difference between data engineering and data science, would you mind explaining with the what is what and why you make a difference?
Hadelin de Ponteves 20:26
Yes, sure. So data engineering is, is more about, you know, processing the data or, or making a pipeline of data. So it can be anything from pre processing, pre processing the data, or, you know, managing high high amounts of data. And you know, I was dealing with big data with a big data of Google. So I was doing a little bit of SQL on he wasn't SQL with SQL, Google. They had their own Big Data Management Program. So Yes, I was doing a lot of a data management but also was doing a lot of data mining. And to to, you know, to derive some insights. And I was building a data tool. So basically, data engineering is the large scope of data science data science has more different science or branches that you can use to manipulate the data. So you can use machine learning, that's a branch of data science to try to make predictions from the data. Also, artificial intelligence is a branch of data science because you leverage the data to make an automated system that completes certain actions to accomplish a certain goal. Pre processing is also part of data science. So yeah, it's almost the same although data engineering is more you know, being an engineer manipulating the data In order to, to do whatever from data Master Data Management or deriving some insights, extracting some insights from the data make sense. Excellent. So why did you decide to leave Who? Well, first of all, the the contract was a had a determined length. So at the end, you know, I didn't want to renew the contract, or I didn't want to, to apply for another position at Google. And that's because, yeah, clearly, I wanted to become an entrepreneur. Even, even at the beginning of my work experience at Google, I knew that it was only going to last the length of the contract the timelines of the contract, and then that I would start my way as an intrapreneur. And that's exactly what happened. At the end of the contract. I, I, I became an entrepreneur and I started these ideas of business I had first my own platform and then the online courses. Did you try
Tim Bourguignon 23:02
other ideas that didn't come to fruition?
Hadelin de Ponteves 23:05
No. Because actually, I was, I was very lucky. Because usually when you start as an entrepreneur, you, you have several ideas, you try several ideas, but the first ideas don't usually work. I'm saying this from the observation of my other entrepreneurs have friends. And and I was very lucky. I'm very grateful that my, we know, yes, my maybe second or third idea was successful. Because, indeed, that best selling course that I released, you know, that this Bible of machine learning, actually, was released six months after I after I, I started as an entrepreneur, which is a you know, a very short period of time before you know, start to see the success of your entrepreneurship journey. So no, Yes, I will. I was probably very lucky or very, I'm very grateful for this. My first step Second idea was the one, you know, that propelled me as an entrepreneur. And that, you know, took me off, I guess
Tim Bourguignon 24:10
you were at the right time, in the right place with the right content. That's exactly yeah. It's also part of the journey.
Hadelin de Ponteves 24:15
Yes, when I do a retrospective, you know, observation of how I got successful. I, I realized that I was very lucky with the time because the time when I released that Bible of missionary machinery eight is it was a perfect time, it was really at the inflection point, the demand and the hype for machine learning. And there was already Of course, a lot of online courses in machine learning already published and a lot of books that as I was telling you, you know, I read all these books, I did all these online, I took all these online courses, so I knew there was still a way to differentiate with something more complete something more complete something more structured, something that provides trainings in both Python and R, and undetected. That's exactly what happened. So not only I differentiate it, but also it wasn't the perfect inflection point when it was released. So that was really a lucky one from
Tim Bourguignon 25:19
take us to blue life. So what's what's the story from from this ecourse intrapreneurship and then shifting to where blue life and how did you create it?
Hadelin de Ponteves 25:29
So basically, I created my first online course in the 25th 2016 end of 2016 years. And then it got so successful that I created many other ones. I developed some kind of a an addiction to for it. So I ended up creating one course per month with my business partner. So yes, every month we were releasing new courses on deep learning artificial intelligence, advanced artificial intelligence chatbots natural language processing blockchain and Many courses today we have more than almost 60 courses published. But then at some point, you know, you, you know, I kept creating all these courses every month. But then at some point you, you know, you start realizing that, okay, maybe you can do something different with your life, and maybe that's what you should do. So, instead of continuing to creating all these online courses, well, I actually, we actually had the idea to not continue to be the business operators of that business, but to become business owners and automate the business. So instead of operating as business operators, well, what we did is we partnered up with some other instructors who would create courses for us. And who would be very happy to do this because we already had more than half a million students on Udemy and therefore, it was a great way for these partners. Partners instructors to penetrate the market on Udemy and get a you know, some kind of a shortcut. So yes, that's what we started doing. We hired a team, to automate the business to partner with instructors and to continue to release a course one course every month. And, and, and still, you know, grow the business and, and because we were not the ones making the courses, while we had much more time for ourselves to do something else. That's how we created blue life. That's how we created blue life, which is very different business, a b2b business, where we would leverage our expertise and experience in artificial intelligence to help other companies innovate with AI and, and you know, maximize profit, gain efficiency, solve problems, improve customer services. Well, any application that you can today do with artificial intelligence. So that's how it was born. And we're very excited because yes, a blue light is, is a is very rich in applications, you know, we realized that AI can be applied to many, many industries. So we have opportunities and clients in a variety of industries. So we get to, to learn about different business knowledges. And we get to, you know, participate in some in a variety of challenges of this 21st century. And that's a very, very rewarding,
Tim Bourguignon 28:30
so you're not just consulting, you're also developing or creating solutions for for those customers. Oh, yes,
Hadelin de Ponteves 28:36
that's right. I am not the one to do this. Because like, I am more, I have more of a director role. Therefore, what my goal is to, you know, expand the business grow the business, and therefore, what we did is we hired a team of very talented experts who developed the solutions.
Tim Bourguignon 28:57
What is the story behind the name Life.
Hadelin de Ponteves 29:01
Okay, so that's pretty simple. When we, and that's a funny one, when we decided to start a new business, you know, to automate the online course business and start a new consulting business. We were actually in Portugal, in a restaurant in front of the seat. And, and so once we decided that we, the first thing we wanted to, to figure out was the name of the business, the name of the company. And, and so we actually I think we stayed for two hours. And the lunch, we were having lunch and for two hours, we were trying to figure out a name. And and so at some point, I think at some point in the conversation, I don't remember if it was me or my business partner Kirill, but one of us asked, okay, what inspires us the most today and, and may I say the water because I really like to water. I really like to see I really like the blue of the ocean and everything. So I say this. And, and, and then he says, Oh, me too. And or maybe that's the other way around. I don't know who said that first, but very given anyway, we agreed that you know, the blue is the blue of the ocean really inspired us. And so that's how Yeah, we simply came with the idea of blue life.
Tim Bourguignon 30:26
That's interesting story. Yeah. Let's go back to you. How do you continue or do you continue to scratch your your technical age and power up a shell and write some Python scripts and, and malloc some data yourself, or did you completely switch to a to a management role or a zero? And don't go too deep anymore in the in the technical side?
Hadelin de Ponteves 30:54
No, no, I yeah, indeed. I I really are more today in management role and commercial role and business rule. However, I still have a passion for, you know, developing AI's. And I don't, I don't code much. Yeah, I don't code much today. However, what I still like doing is a stay ahead of the game in AI by following with state of the art following the state of the art of artificial intelligence. So what I what I still do today is read the research papers. I love doing that I love reading about the new models that are invented, for example, and fascinated about this new breakthrough in, in natural language processing with, you know, Bert, developed by Google. So this and also I'm really passionate about the reinforcement learning branch of artificial intelligence. And yes, so I keep I keep it I stay ahead of the game by reading articles, research papers, blogs, about the latest advancements in, in artificial intelligence and data science. But however, yes, it's true that, I don't know, I don't, I don't code as much as before I code sometimes because we actually, you know, one service of blue life is to offer corporate trainings. So we go into companies and we give corporate trainings on machine learning, artificial intelligence. And of course, inside the corporate trainings are lots of practical activities. So I actually code the practical activities with you know, the the team of employees that I work to whom I train artificial intelligence. And so therefore, to prepare these corporate trainings, and actually prepare some case studies where I code some artificial intelligence to solve the case studies. So yeah, I still I still code from time to time but not as much as before, obviously, and And the way I really stay in touch with AI, with the state of the art of artificial intelligence is mostly through research papers and books. Do you miss it? Well, the problem solving, I still do it. Because, you know, when, when I try to get new projects or clients, well, my first my first reflex is to understand their problem, and then brainstorm how to solve them. Because the way we sell artificial intelligence is by first, you know, brainstorming with the client, and figure out if AI can help them, you know, AI can does not necessarily help everyone. So, at first what I do is I make sure that, yes, we can bring them some added value with artificial intelligence, by understanding what problem they have to solve what they need, and how we can solve it. So I still do a lot of problem solving, although I don't do it You know, by coding. So I guess your question was, do I miss? You know, solving problems directly by coding them? Well, yes and no. Yes. Because Okay, there was there was very fun to do. But no, because I actually really enjoy my new role as a businessman and as a, as a management as a manager and as a business developer, because that's, I think it's very, very exciting. I really, I really have a passion for my different businesses, including blue light. And it's really a huge excitement to to develop it and to and to grow it. So because I have this huge excitement, while the previous excitement of solving problems with coding kind of, lets lets itself being forgotten, you know, so yes, when I when I read, think about it, I kind of miss it, but then since I'm more into the My new role and I really assume and and identify myself as my new role Well no, I don't actually think too much of it what's
Tim Bourguignon 35:09
what's next for you? Where do you see your business going? Where do you see AI and data science going
Hadelin de Ponteves 35:19
my, my vision, my really my core goal and objective is to grow blue light and grow my personal brand because I have developed very, very good personal brand with all these online courses are successful. So, I work a lot on you know, you know, be becoming like a top influencer in artificial intelligence. And also I would like to be one of the guys that will have a positive impact on the world thanks to AI. So, yes, I work both very much on my personal brand development and on blue light. And so, for example, for my personal brand development, I will I spent the whole last year writing a book and book that was just published actually published last month, which is called the AI Crash Course. And yes, this is a book on of course on artificial intelligence. And in this book, not only I explained the whole theory of artificial intelligence with many practical case studies, but also in this book, I, I incentive a, I actually highlight my purpose, which is to incentivize people to do better things in this world thanks to artificial intelligence. So, so yes, my my purpose I would say today is to you know, spread a high quality education on artificial intelligence, help businesses, solve their problems, innovate the right way, and, and, you know, make a puzzle impact in this world, while at the same time through my personal brand and influence, in some incentivize people to do the same, no to do artificial intelligence for good to convince them to, you know, work on artificial intelligence by reish, reliving them that it's not as hard as it may seem, and by incentivizing them, convincing them that they can do it, they can use it they can apply it for great applications. When we think about artificial intelligence and and data processing and the sciences hope we look at the the news
Tim Bourguignon 37:41
of the world, it's often bad news that's come up. It's examples of our Facebook screamer with our privacy. It's example of data, data leaks that that leads to well Lack of privacy again, it's people misusing our data. And it's always all those all those news that come up. Do you have examples of the other way around of fields or applications? Where this use of data and really make a big impact on the world or positive impact on the world?
Hadelin de Ponteves 38:23
Yes, of course. Well, the first thing that comes to mind is that healthcare with artificial intelligence, we can now detect, you know, heart conditions or, or, you know, cancer signs much more rapidly and efficiently then, then doctors, actually doctors use AI systems to assist them in their, you know, tumor detections, because they're very, they're much more effective. So yes, that's the first. That's first example of how AI really can help us save more lives. Then also, another example is How AI optimizes, you know, energy distribution and which is actually used for renewable energy. So yes, AI has a positive impact on like the fastest growing industries like renewable energy, then AI is being used, of course in an in transportation. And the goal of the industry later leaders and transportation is of course, to leverage artificial intelligence to reduce the number of accidents with a self driving car. So I know we're not there yet. But at least that's what they plan to do with it. So yes, ai ai can have some really beneficial impact in every industry. And the and the reason why we hear all those negative things about artificial from the media is I would say because of the media itself. You know, the way for the media to, to get the the focus of the listeners and the readers or the, you know, the people who watches the Western media is by spreading negative things because negative things is how we, how we get attention. You know, it's not if the media were was saying always positive things, this would be boring. And this people wouldn't get excited. So the way the media spread their message is by using negativity. And that's how artificial intelligence is the same as a, you know, an evil by a lot of people because of the media.
Tim Bourguignon 40:42
That's very true. That's very true. Do you have anything to say about about biases that we might be building into our AI systems? I read the story about the prison system in the US. That is partially Driven by AI now at least the the judges are using some some AI tools to assess what the what the sentence should be. And apparently the system learned from historical data. And since the historical data has some biases in itself and racism, the AI system of the expert system is now biased against against people of color, for instance, and this is something that was just not not wished or imagined by anyone. But by using these expert systems and, and thinking that they are going to answer all questions, but feeding it with a bias data, we, we kind of lie to ourselves. And so this is one of the of the critic I see often about AI systems. It's Yes, the the algorithm is fine. The mathematics is right. But what about the data we feed it?
Hadelin de Ponteves 41:58
Yeah, so that's an example. where the problem is indeed in the data? So, yes, so absolutely in the data, you can find me biases. However, there are some techniques. And I'm surprised that they weren't used but some techniques of evaluating the, the, the predictive model of the AI through evaluation sets. So, you after you train your artificial intelligence on, as you say, historical data, you want to test it out on actual data, you know, actual data at the present time. So I guess the issue in that was that the evaluations were not done properly, and test sets were were not, could not identify the biases that were found in their historical data. So that's why it's very important whenever an AI is built and trained on data from the past to, you know, to confirm it on new data on new data that we call test data or evaluate data evaluation data.
Tim Bourguignon 43:10
I hope so that people like you are really, really having an eye on this and
Hadelin de Ponteves 43:16
doing the right thing. We're very careful. Yes, yes, we were very careful with we, you know, we reassess the AI several times with a new test set. And, and we always make sure that that no, bias was left inside the data. If
Tim Bourguignon 43:37
you could travel in time and and go back to, to to your room. When you were starting your studies. What advice would you give yourself?
Hadelin de Ponteves 43:48
Good question. I'm actually pretty satisfied of the different decisions I made. So I would say, well, the advice I would say is to trust yourself. If That's what I did. So it's not really an advice of giving because that's something that I did naturally. But yeah, trust trust yourself. Trust that it's going to be okay. and work hard. But again that that's what I did. So I guess you will you would expect me to say something that I didn't do correctly. Not necessarily
Tim Bourguignon 44:19
No, no. It's is that the advice you would give somebody someone else to someone else is different? Because
Hadelin de Ponteves 44:25
yes, then I know what to tell him. But to myself. Yeah, to someone else. Yeah, I would say definitely to. Well, first of all, to trust oneself in in what they're going to achieve in the future. I would say also to, to, to work hard to work smart. too. Identify your purpose, because I kill you had a purpose myself. But yeah, I would say that it's clearly important to identify a purpose because the most important thing that I needed during my entrepreneurship life was inspiration. I had it most of the time, but I lost it at some point. And I can tell you when you lose it, you know, things go wrong. You You don't believe anymore. You don't. You don't. You're starting to doubt. And the only thing that will keep you inspired is to stay focused on your purpose, to stay focused on your goal, you know, so the best advice I would give is to identify this purpose to go so that whenever you lose inspiration or your intrapreneurship journey, which will happen, you can get back to that inspiration by remembering what is your purpose? That is
Tim Bourguignon 45:54
very deep, thank you very much. Well. So where could the listeners Start or continue the discussion with you if they wanted to, where would be the appropriate place? Well, the social
Hadelin de Ponteves 46:06
media I spent my most time, the most of my time sorry, is on LinkedIn, and on Instagram. So LinkedIn because I also have a passion for traveling. So I kind of share it on Instagram. And LinkedIn is, of course, my professional network, where I, you know, post the latest news about AI or my latest product. For example, my last post was about this book I just published. And this created a good way that excitement. So yes, the best way to basically follow my my professional journey and my entrepreneurship journey is on LinkedIn. And then for more my personal life, also my professional journey, Instagram. So my LinkedIn is Atlanta controverse. I guess you will share that on on the page and My name is Helen to pee I wear I guess you will also share that. But that's the to me
Tim Bourguignon 47:07
again and what's next for you in the in the upcoming month, something to plug in something happening that you want to advertise.
Hadelin de Ponteves 47:14
So I'll do several book signings all over the world. The next one will be in Seattle in March, actually will be the beginning of April, then it will be in Germany. Then it will be in in France. And in the UK. Also, my business partner hosts his a conference where I kind of participate heavily because each time I give a one or two we're big workshops. And this conference is called Diaz go data science go. And actually, there will actually be two conferences this year, one in Berlin and one in California. The one in Berlin will be happening in May, and then one in Calgary. will happen in October. So David Simon's good at calm. And then, you know, I'll be traveling everywhere to meet clients to meet new opportunities for blue light trying to grow blue light. And, and yes, so I guess they will know where I'll be from the social media and, and if they want to catch up over a coffee because I always like to, to catch up with a for example students whenever we are in a new place in a new country, we always catch up with students for example, with the trip in Europe, where in every new European country we would invite 10 to 20 people for dinner, you know, and these people were of course people from our network or people from our students viewing so I that's something I really like to do you know, get a simple coffee or get a have a dinner, and discuss data science or AI over dinner with us today. With the people who follow us, we always like to do. And yes. So it's usually very easy to find me. It's not like I'm a very passive on the social media. So yeah, I guess they will have no issue to find me.
Tim Bourguignon 49:17
Okay, so listeners, you know what you have to do? Yes.
Hadelin de Ponteves 49:21
Tim Bourguignon 49:22
Thank you very much. It's been a blast. Oh, yes,
Hadelin de Ponteves 49:25
indeed, this was such a great discussion. I was very excited. And, you know, it's always by talking about childhood and, and all memories that, you know, you get very inspired to, to, to continue in life. Amen.
Tim Bourguignon 49:43
And this has been another episode of developer's journey. We will see each other next week, bye. All right, this is Tim from a different time and space with a few comments to make. First, get the most of these developer's journey by subscribing to the podcast with the app of your choice, and get the new episodes automagically, right when the air. The podcast is available on all major platforms. Then, visit our website to find the show notes with all the links mentioned by our guests, the advices they gave us, their book, references and so on. And while you're there, use the comments to continue the discussion with our guests or with me, or reach out on Twitter or LinkedIn. Then a big big THANK YOU to the generous Patreon donors that help me pay the hosting bills. If you have a few coins to spare, please consider a small monthly donation. Every pledge, however small counts. Finally, please do someone a favor, tell them about the show today and help them on their journey.