00:02:34
It was the second edition of a book on cloud, and really helping, that was, I would say, half technical, half business, and really helping customers and companies to understand how to move to cloud technology. And because cloud is something which, you know, everybody speaks about, you barely know that there is server somewhere. But the world as it is today is much more complex than it used to be, and it's not just how we're going to move everything to the cloud. So it's a kind of a multi-cloud and hybrid cloud environment. So how do you serve all those kinds of complexities as well, how to work on multiple clouds, multiple vendors, local as well, when you are running local servers. So how to make sure that you're doing the right choice from a security, maintainability, and the whole, you know, serviceability perspective.
00:03:25
That's super interesting, because usually in our podcast, we are covering a lot of artificial intelligence, but for a lot of episodes right now, we didn't cover actually cloud. So can you tell us more, for example, about like how cloud, like what is actually the future of the cloud? Like, because we know, for example, from the AI perspective, we are like foreseeing like AI agents, et cetera. But from, for the clouds. There needs to be a cloud. There needs to be a lot of development also to absorb all of the emerging technologies. So what is the future of the cloud currently? Well, difficult, difficult to say, you know, I cannot really foresee the future, but I think that the big trends that we're seeing, well, AI and generative AI particularly is definitely very cloud-based, software as a service, you know, SaaS-based.
00:04:16
So if you look at what OpenAI is doing, Anthropic, Microsoft, Amazon, all those big companies now there. They're really running their own infrastructure and trying to basically get you on their infrastructure. So I think that that's where a lot of people are going, just because, well, it's cost-effective. Okay. If you look at how much it would cost to run a local environment, it's just mind-boggling. Most people don't go into something which is not cloud-based. Now what we are seeing and what I'm seeing as well is more and more concerns about security. And that means that from a security and privacy perspective, a lot of people want to or think about how could I bring some of that compute power in-house or locally at least.
00:05:06
And so, what we're starting to see, and it's very interesting as well when we're seeing what Meta is doing, for example, on that front or Mistrial in Europe, where you can run on open source local models, including models right on your phone, on your PC. You know? So the latest announcement as well from small language model running on just normal CPUs instead of big GPUs. And so I think it's a trend that we'll see going where you can run your own model while benefiting from the larger as well in the cloud. And I think that's an overall trend where you want to mix local compute power, external compute power, and so creating your own hybrid environment. And I think that's where, if I were to place.
00:05:53
I think it's probably where people will be looking into going more and more, keeping some data locally, benefiting from large models, but running as well locally when we can. So one more, can you repeat what exactly the hybrid environment means for the end users, for our audience who is not that much technical in this space? So what are you excited for? Yeah, that's true. So in a nutshell, if you go. And let's go, let's take ChatGPT for example, okay, that's, that's, you know, something that most people understand now. So when you run your, your prompt on ChatGPT, if you're using just ChatGPT. com, so you will run completely on what we call the public cloud. So which will be somewhere, you know, we don't know where the servers are really.
00:06:41
So probably most, most recently in the US. So this is a public cloud, something with your Gmail, with most of the systems that, you know, companies or individuals are using. So there, there's somewhere. Now, on the other side. You've got organizations that have a data center and they are running their own environment. So you know, whether you're a small company or a large company, so you run your own environment and we can call that in some respects, a private cloud. So this is your own environment. Now, the moment you want to make a bridge between what runs in the public cloud and what runs on your private cloud, then you're creating something which is a mix of both. And we call that hybrid.
00:07:18
Of course, we don't have really time to go into more details from a technical perspective, but it's really making the use of private servers, public servers, generally on a multi-to what we call the multi-tenant environment, which is shared environment like Open AI. And so, but we're making one single cloud from a usage perspective and from an end-user perspective, for example, I can run my, or some, some of my workloads locally, some of my workloads in a public cloud. And from an end-user perspective, I don't see the difference. Okay. It's just, it's just service. And that's, that's what we call a hybrid cloud. It's a mix between public and private resources. And everything, like the main benefit of this is the security, that actually you can store your private data and you know where your private data are, because I think that lots of also like end users on the internet, they don't really, to some extent, sometimes care about like if it's like on public or private cloud, to some extent.
00:08:21
And I know that usually, like when I'm speaking with really IT people, like developers, et cetera, they really care a lot. But for example the normal users who use the internet, for example I don't know, students in the master degree of international relations, like they don't usually dig deeper into like where their data are stored, et cetera. Yeah. And it's true. And I think that there is a, an evolution actually in the, in the privacy view. Which most people, you know, I would, I would say they don't care. I think they don't realize. Yeah. And you know, and something that I keep telling people, you know, just if you take a picture with your phone, whether it's an Android phone or an Apple phone, whatever, generally those pictures go immediately in the cloud, okay.
00:09:11
Whether it's iCloud or Google Store, or whatsoever, and then it goes directly to the cloud, and nobody cares. But the fact is that it is indexed. AI will probably, probably labeled those, those pictures automatically. But who cares? That's a picture. Now, when it comes to confidential information, you know, customer records or your bank account information, those kinds of things, why you start to say, 'Hey, I don't want to share that with a lot of people.' And most organizations think about, okay, I want to keep privacy on those confidential data and keep them locally in my private environment, but benefit as well from services that are external. And as I said, in the case of AI, for example, AI requires a lot of computing power.
00:09:54
And you know, if you were to purchase a server that would probably cost hundreds of thousands of dollars where most people can't, cannot afford, well, you, I can keep my data local and then use compute power from a public cloud. But yeah, we need to make a, to make a decision where to store our data and how it's used. Yeah. We should be aware. And also to spread awareness. That's like the people also who are currently watching us. Yeah. I should also like think about, you know, like just puzzled, like where all of your data are going and like to be mindful about your data. There is also like big, for example, mainstream between the different countries in the European Union, for example, that for example, Portuguese people, they totally they are decreasing their digital activity that they don't want to be on internet.
00:10:41
And like, for example, Czech people, they are more on internet. So it's also like quite cultural that will have consequences. Yeah. It's also in the future because we are going more and more digital and into the digital sphere. I also like how you are usually like presenting yourself on social media. It's like you are like a disruptor in, in the technology that you like to really like think deeper into different topics. Like for example, what you have published recently on LinkedIn about artificial intelligence and the Roman empire. Yeah. How like is it that you. Yeah. So I do read a lot of the books, but I don't really like to read a lot. And then you really connect the different dots, like how is it now in the modern history and you connected to the, how is it right now in the modern times and you connected to the history or how, how did you come up with this, for example, article.
00:11:37
That was not an idea of mine. So you know, we, we need you to, to, to credit Romans for that. There was an article that I read and actually that I quote in in my post. But but I think that. The thing is that there are two things about, you know, how I see technology. The first thing is, as I said, I'm an engineer by study. And, you know, engineers, what we are taught at university is how to understand how things work. And so it's not just, you know, taking, I would say, a very logical, not logical, but a theoretical view. So we are really hands-on. So you want to understand, you know, engines work, how a computer works, how programming works and stuff like that.
00:12:17
So I'm always curious about understanding and going a little bit deeper than just, you know, what happens below the surface. As I used to, one of my first cars, when I purchased my first car, it did not have a heating system. So the heater was not functioning. And when you live in France and in winter, it can become very, very, very complicated. And so I decided just to tear it out. So I said, I'm going to fix that thing. And I basically dismantled the car. And when I re-mantled it. So I rebuilt everything. I had a couple of spare parts in my hands. And I don't know where these things are going. At the end of it, the car was working. So it probably was useless. And I think it's the same thing with AI.
00:12:56
You know, I wrote my first book on, what, 2018, something like that, 2018, 2019. When very few people were talking about AI. You know, ChatGPT was not out. But I really wanted to understand how these things were working. So I said I used to run Azure for Microsoft for a couple of years. And Microsoft were really starting. AWS was starting as well, you know, in this AI system. So I need to dig deeper. That was not something that I studied at university. So I said I want to dig deeper. And so I dug deeper. I tried to understand. I tried to make my tests. I run models and those kind of things. And, of course, when you look at what OpenAI has done and what large language models have done, you start questioning how these Transformers stuff work.
00:13:35
And so, of course, you immediately come up. And this Roman Empire stuff was very interesting because basically every technology, every group of invention of people. Go up and it's like the hype cycle. So we go up. Everybody's interested in, you know, it grows and grows and grows. And at one moment it collapses. And what can we learn from those collapse? And I think that's what we are seeing with large language models. And I'm not and I'm honest. I'm not a complete expert about that. I understand how those things work quite deeply. But I know that a lot of people are looking into other dimensions of AI and natural language processing. But from what I read, and from what I studied, it feels like we will hit a limit one point.
00:14:21
And so something will come up and say after Roman empires, we had, you know, Christianity and middle age and blah, blah, blah. And the whole thing. And I'm not an historian, but basically, history, history of mankind is always there is somebody who goes up and then goes down and then somebody comes up. And and if you look at it, it's the same thing. We had client-server architecture in the 90s and then client-server, which completely dropped because Internet came in. And so, so we had anti-architecture, multi-tier architecture. And so, and technology keeps reinventing themselves on a constant basis. And so, so that was, and it was quite, you know, fascinating by the fact that, yeah, so maybe there are some parallels between what we're seeing in AI and what the Romans did.
00:15:02
Or we could have I could have done with the Greek that probably could have worked the same way. Yes. I also heard this actually this week, which is quite coincidence. And it's that the history is repeating. It's just the evolution of technology. We are that even when we were really like at the beginning of the evolution, we needed always to have a water, food, shelter and the fire. And in this case, like the fire is the technology. It's the innovation. It was the most innovative thing that we had. And now the fire is the Internet AI and it's like turning. We don't know. We don't have no longer like in front of our apartments, a fireplace place, right? Like we have a light, we have a bolt. We are using.
00:15:45
The AI for the search for like how we can get the information faster to us. So it's always like in the history, the similar kind of phenomenon, which is like changing the beginning of the innovation, which is fire. So I found it like super fascinating. And it also like why this article also catched my interest because I found actually I didn't find at that time and I was reading the coincidence with the fire. It was just right now. When you are. Repeating. It. But I have also like because you are a lot, you are a lot doing right now with artificial intelligence, artificial solutions. So you are a lot in like technical, you're engineer. I have a question. What was the most innovative stuff that you have seen recently?
00:16:34
Something that I know that you've seen a lot of stuff also during your career with the Microsoft, et cetera, but something that catched your interest, something that you were like, OK, yeah, this can, for example, this can, for example, save me a lot of time during my daily life professionally or personally. What was the most most innovative stuff that you have recently come across? Well, that's that's a that's a difficult question. I think that's I'm amazed at everything that comes out almost on a daily basis. Well, not daily, but almost weekly basis. I think recently what I I completely was was really. You know, astonished about was what Claude, for example, has been doing. So the I thought that the evolution of three point five sonnet and how it works now.
00:17:27
And and I find it fascinating because when I started using like all of us, you know, generative I late twenty twenty two. So it was kind of magic and then it becomes less magic and then and then, OK, says, yeah, fine. So you ask a question to chat GPT or to Claude or whoever. And then you got an answer. And then the answer started to be a little bit less, you know, imaginative or or you started to get bored somehow. Yes. Change BT for was good down the whole thing. But then recently when I and I think that I skipped that update from Claude and then suddenly we had this split screen where you ask the question, you have that that that text that Claude writes and then you can interact with.
00:18:11
I want to change that paragraph. And then and then you start having really, really a conversation with and, you know, it's a machine, but it becomes completely pervasive. And I scratch my head and say, 'Gosh, I'm not sure whether it's a machine or, you know, somebody who is typing at the back.' And I think that it was that was kind of interesting in how you interact with machines. And I think that overall, to me, what I'm really fascinated about is this ability to interact with machines in a in a. Seamless manner, you know, and as I told you, so I wrote my first book on AI was six, seven years ago. And at that time, even when I was doing some testing, for example, on text-to-speech or speech-to-text, it was kind of yeah, it was good, but it was not perfect.
00:19:01
Now, you ask questions to chat GPT, just ask the question, you talk, he talks back and then you're starting to have a conversation. And I often go for a walk in the forest, you know, and I put my headset on and I start talking with. If somebody crosses me, says this guy's crazy, he's talking, he's talking alone. But I'm fascinating because it to me, from a creative standpoint, it's kind of you ask questions to somebody who has access to the whole world of information. And and you have really deep conversations. And it's to me, it's fascinating in the way that it unleashes your own creativity. I have to try this. And I think that our audience, whoever is watching right now, just go out.
00:19:44
I think that this is also like pretty much like amazing, like a mental hygiene that's like people can because I also think that sometimes like psychologists have that much demand because like people don't have with whom to openly speak about the things. And if this chat GPT can actually like do this kind of job, where people can just speak openly about their questions, like, of course, like meaning of life. You should bear in mind that it's still a machine there. But then, like, if you will reach the point that OK, like a machine cannot solve my issues, then you go to the psychologist. But I also came across that a lot of people go to psychologists just because that they cannot like openly speak about all of their thoughts.
00:20:32
And so GPT can like open the creative mindset. Yeah. And, you know, one thing that I like actually is asking questions to chat GPT or Claude or whoever. Or he asked me question, you know, it's about a, you know, and sometimes God, OK, I got this problem or let's say, well, what was when was that two, three days ago? I said, OK, I was contacted by a journalist who wanted me to write an article on a specific topic. And then I scratched my ass. OK, generally what I do is I go to Chat GPT. You know, I've got this problem that I don't. Can you help me? And I always and I went exactly did that. But I did not type. I talked. I need to write this article on, you know, for that particular magazine.
00:21:13
So I have a couple of. So here are my top ideas on AI and digital transformation. But what do you want to be more insightful for, to help you write that article for this particular audience? And then the AI starts to ask questions where I need to understand, you know, those things and those things and those things. And then you said, OK, so let me let me provide you this answer. And then we start having a conversation. So it's not just about the AI providing an answer. It's an AI asking insightful questions. That forces you to think and to provide more information so that you really have some. And I will I will say somebody because I feel sometimes it's somebody next to me that will help me think and not just provide an answer.
00:21:58
And that's I think that to me, that's fascinating because that the future will be very interesting. Yes, for sure. And I also like this that, like, actually, you let the AI to ask you a question. I never thought about this for me. It's like input output. It's still the mindset. It's still the mindset of the Google, you know, the Googling that I need to provide the context to the AI so that it gives me the result. But also like this mindset will evolve with us as humans that like it's not just the Googling, like it's not chat GPT, right? It's a you can actually have like different kind of narrative if you just switch your mindset about it. So this is also like super interesting how we can change as a human being towards the technology because it's out there.
00:22:43
It's a tool for you, and you should leverage it to its fullest potential. You speak a lot about ChatGPT. What about others? Do you also use like Copilot or other models, or chat GPT is your go-to? Well, I think I'm using Cloud more and more. I think I've been one of the first users of Anthropic Cloud a long time ago. I think that it's very creative. And I would probably juggle between Cloud and ChatGPT. Okay. So what I've done with ChatGPT that I didn't do with Cloud is developing my own agents as well. So, I got my own GPTs when I need to do specific things so that I basically prep them to really answer some specific things. For example, I was telling you that I was teaching in universities.
00:23:31
And generally, when I need to grade assignments, I create my own teacher, basically that will grade the assignments for my students. And of course, I prep them. This is what this is what I'm looking for. This is how we're going to grade those assignments, those kind of things. Okay. So, I'm adding a lot of examples as well on what to look for. And he's doing a fantastic job. So, saving me hours and hours. So, yeah, but I would say I'm juggling between Cloud and ChatGPT mostly. And that's, yeah, that's the two big ones. I tried Gemini, Copilot. I'm not a big fan. And then when I need, I'm really using actually the OpenAI services on Azure where it's, you know, it's more private. And I can develop as well flows.
00:24:14
With Microsoft Power Automate. So where you can interact with reboots, robotics process automation with some artificial intelligence on it. So that's an activity as well that I'm developing. And what would you advise to the people who don't yet use AI? Like, I mean, like nowadays, like everybody is speaking AI, but there is still like a proportion of the people who are like receiving this information, but not really like integrating it in the practice. So what would you advise to these people, how to start with artificial intelligence, how to really stand towards this position of, because is it actually necessary to jump on this board with the artificial intelligence? So I think, I think it's like, you know, learning how to read or how to count. I think that it's, it's, let's be honest.
00:25:06
I don't think that AI will disappear anytime soon. It will evolve. It will change over time. But there is something that, that people. People need to understand is how to interact with those machines. And so the so-called prompt engineering, because asking questions actually to, to an artificial intelligence and to regenerative AI particularly is a kind of an art and a science at the same time. So you need to ask insightful questions. If you want to get the right answers, we'll, we know what, the more you use it, the, the, you understand that when you ask good questions with details and guidance, then you get a very interesting and detailed answer. And therefore the thing that I would really advise people is, you know, peak chat GPT, cloud, Gemini, Google Gemini, whatever.
00:25:50
I think that everyone or every, every tool is good in a sense. Now, I've got a preference for ChatGPT because they're the first ones and Cloud because they're more creative in both environments. And I'm thinking ChatGPT for an Open AI, the Open AI website, they have a piece of it where you can learn how to prompt. Okay. And then we provide guidance. And of course, you can find that on YouTube as well, or many other websites. But it's really around don't go and chat and start typing questions. Okay. Because you, you may be, I would say, mesmerized at the being say, wow, that's, that's magic. But then very quickly you will find some weird things happening and you will, you will somehow be disappointed.
00:26:29
And I think that you need to go through a little bit of learning and, and, and I love the Open AI as the, as a platform that really teaches you or gives you hints on how to, you know, prompt. Okay. How to enhance quality of your prompts. And so spends, you know, it's, that's not a long thing. It's probably half an hour to 45 minutes. Just spend that amount of time just to understand how to, to query the machine. And once you've done that, I think it's a, it's a, it's a no return journey. So you, you will go and try to, and honestly, that, that's something I say to people as well. You know, what, what could happen? Well, not much, except that you will discover something that you haven't thought about.
00:27:12
And it goes to, I was, I was showing that the other day to a couple of friends says I was alone at home. So my wife was not there. And I say, I need to cook something. And said, I don't have a clue of what I'm, what I'm, what I'm cooking. So I took my phone. I really opened the fridge, took a picture of the fridge and asked Chaji PT, what can I do with that? And it provided me a couple of recipes. And I say, well, that's pretty cool. And I don't, I don't remember what it was, but it was an Asian food, something. He says, 'Well, that's pretty cool. I've never cooked that. Let's do that.' It was not perfect, but I had a good time.
00:27:46
And that's something that, man, it's just take a, take a picture. And the other day, same thing. So we are at an airport with my wife and there was just a plane parked. So we're having a coffee and there was a plane parked. And we could not really figure out what plane it was. You know, we don't know where it's coming from. And so we took the pic, we took a picture of it and asked Chaji PT, 'Where is this plane coming from?' He says, 'Well, it's a Georgian plane.' It's coming from Georgia. He provided me the whole thing. He said, man, it's just, gosh, I got basically anything. I can take a picture of anything and get answers. I can listen to a conversation.
00:28:21
Same thing the other day, a friend of mine says, I need to, you know, to program that CCTV camera. It's coming from China. And the only guide that I have is in Chinese. He says, yeah, bring it to me. I got a good friend who speaks Chinese. And I uploaded that to Chaji PT. He says, you know, can you translate that into English? Boom, done. Man, this is, this is a fantastic. Yeah, that's really amazing. Like also the globalization that also we will be able to more communicate together, as you said, about the Chinese, et cetera. Like, because like, for example, I was in November last year. So one year ago I was in Japan and I was super surprised that they don't really speak English.
00:28:58
And they have a lot of like translators that like you go to the shop and they use the translators, not, not even the phones, but the translators. And they are like speaking Japanese in it. And it translates. Here like AI is like super helpful. Like they didn't have really like powered by AI yet, but I'm sure that somebody is already working on this solution. We will be able to more like communicate among each other, which can even help the politics, et cetera, because I worked in a European commission for some time and the translation unit there, it's just immense. And it also consumes a lot of expenses. Then like a. Yeah. During the translation that they can just like, you know, misuse some board, it causes misunderstandings.
00:29:46
And I think that this is really like, as you said, like amazing times to live in. Yeah. I was actually talking about Japan. So I was in Vietnam earlier this year and the same thing. So took a cat taxi. Those guys don't speak English. They don't speak French. They don't speak Vietnamese, but the taxi driver had his phone. And so it started speaking Vietnamese. The thing started to translate into French. I spoke French and he was French. And that's, that's crazy, but that's funny. Yeah. The conversation at the end. Yeah. That's, that's wonderful. And you currently, I see also like that you said that you work, you work a lot with AI and providing AI solutions to the clients. Are clients of yours usually corporations, small, medium businesses, or is it like individuals?
00:30:34
Who is your client? It's mostly, mostly, mostly large organizations. Yeah. And you are providing them with the solutions to their current processes, IT processes, how they can leverage the AI or? Yeah. So up until recently, it was really just providing consulting services. And now I'm building with a, with a partner, we're building a company that we provide solutions that are really, but, but with the twist, because I truly believe in, in I'd say usability and the use cases. So we look into what do you want to do with it? You know, something, for example, you were, you were mentioning co-pilot. Co-pilot is, is, is great because it's integrated with office and, and cool, but at the end of the day is, is what do you want to achieve with artificial intelligence?
00:31:20
Yeah, of course you can create, as we shared, yeah, you can take a picture, get some information, create some text. Great. That's fine. But in a large organization, it's really what do you want to achieve? Do you want to increase efficiency? Do you want to streamline some processes? Do you want to enhance your customer service department, for example? So, we need to look into what is your use case that you want to to to enhance, what is your environment that you have and how we can bring some artificial intelligence, process automation, so that we can reach what you want to achieve. So it's really looking into what is the outcome. So it's taking an outcome-based approach. So what do you want to achieve?
00:32:00
And then we can build a solution and help you build a solution for you. So that's, something I'm doing more and more. That's amazing. We actually had a discussion about it. It's also on Monday with Dr. Sabba. She's an owner of a business, which is called Designing Schools, and she uses design thinking; and she's teaching, for example, educational institutions about the design thinking and how to use it for the artificial intelligence, which is exactly like asking first, what do you want to do? What do you need it for? Because it's a nowadays like we are super overwhelmed by all of the softwares that are out there. And to pick one it's really just like you can just close your eyes and pick one if you don't know what you exactly want to do but if you know what you want to do
00:32:45
then it's much easier or harder actually because then like if you don't find what you are looking for Yeah it's true But at the same time I think that you know you need to, to think into the problem one of the problems that I'm seeing a lot with people talking about AI is as I said they think it's magic They think it's going to solve every problem that they have. But what I tell people is that it's, AI is just a tool. Yes, it's a very powerful tool. That you have in your toolbox, but it's a tool and therefore you need to think about how you're going to use that tool. So, and I, you know, there's a joke that says this: if you had only one hammer, then all your problems become, become nails actually.
00:33:22
Um, and, and, and I'm seeing a lot of people say, okay, well, AI is doing everything. So let's bring AI and do everything; but, unfortunately, no, it doesn't do everything. There are some things that it does very, very well, and others that it does not very well. So you need to understand what you want. You want to achieve and it's not really around what you want to do with it is what is the outcome. Okay. What do you, what's the problem you want to fix in your company and your problem? Maybe I've got too many complaints for customers. I need to increase quality of my products. Great. So we're going to focus on quality and you were talking about design thinking, and I love design thinking in the fact that we need to think with the customers in mind.
00:34:01
It's not just about designing a product. It's designing some, something that will be useful for customers. And I think that's. To me, AI, the way that I'm approaching AI is exactly that-of the usage I will put forward. So it solves the problem that I have as an individual or as a corporation. And I think that's the same thing. Yeah. That's a; I also like how it's-how it makes process faster because I heard, for example, from one designer, a thing that-what was usually an outcome before the AI is currently like, for example, like when you, when you're are a designer, right? Like you are first like sketching for your client and like you show, for example, something on the paper, like, this is what I have in mind and the client either say, yes, I like it or no, bring me another one.
00:34:50
But with AI, you are already like bringing the process to another level because you don't need to spend like days with this, like sketching, you ask AI to sketch it for you. So you've already spent these two, three days of sketching to actually just like send the sketch built by AI. You, of course, like you check it. Like if this is according to the client expectations and your expectation and you send it to the client, like, yeah, this is something what I have in mind. And the client again says either, yes, I love it or no, go for another one. And then if clients agrees, then you are doing the tweaks by the human check and like working with the different softwares, like Adobe Creative Suite and other stuff to make it like more enhanced.
00:35:36
It makes all of the processes. It's made much faster, it seems to me as well. Yeah, no, that's and that's that speed. Speed is definitely something. But to me, you know, and that's something I was sharing actually this afternoon is speed is something that I can bring. But to me, one thing that it brings is some it's it's solutions that you may not thought about because the way that it arrives to a specific solution is a kind of and again, it's not magical because there is a lot. If you take LLM, so the transformers work a specific way, but sometimes you'll be surprised sometimes by the answers. And, you know, and I always refer to to the feedback that so Lee Sedol, who was the Go master, who was beaten by AlphaGo, so the DeepMind software, and Sedol at one point says, 'I thought I was about to win that that game against the machine because he did a move that was stupid.
00:36:33
OK, that was really a stupid move.' And then a couple of moves down the line, I suddenly understood that, gosh, it was brilliant. I thought it was stupid, but it was brilliant. But the problem is that nobody, no human being would have made this move. It was the first time I saw it. That's that's when and he said that's when I understood that I was not against another human being or a very intelligent, very, very well-experienced Go player. I was against something which was different. And I think that if you approach and I think I kept that in my mind, because if you approach. If you approach AI as another someone, well, OK, you will get good answers. But if you approach as being something else, an alien kind of thing, then you will be surprised.
00:37:18
And to me, that's the approach. And IÂ understand again, because I somehow understand how these things work from the internal functioning. I know they don't think like a human being thinks. OK, of course. But I want to be surprised. And I'm honestly. I'm almost always surprised by, you know, the creativity, the ideas that I would probably never thought about. And so prepare to be surprised. And, you know, the question you were asking about how we can start, which is just start and you will be surprised for sure. Just start and you will be surprised. I love it. Also, I need to ask about Microsoft, because you spent quite a lot of years there regarding the technology. Microsoft is really a big player.
00:38:06
I have a I have a question regarding what you have learned from like working in Microsoft, like what, for example, other companies don't really apply. My question is leading because like on Monday, Dr. Saba, like she actually worked in Apple for a long time. I asked her the similar question and she answered that the culture in Apple is something different than, for example, in other companies. And I don't want to spoil right now, like what she answered me. But the same. But the same question I want to ask you, like Microsoft has some mindset when approaching new innovations and the technologies that you, for example, didn't see in other companies. Of course, like you are bringing it with you. But yeah, well, I'm biased, but I think that that's a conversation I had a long time ago.
00:38:55
My father, my brother used to work at Oracle. And so we had a, we had interesting because I was at Microsoft. He was at Oracle, you know, kind of, you know, foes and enemies. And we had interesting conversation because I thought when I was talking to him that I could never have worked at Oracle and he probably couldn't never had worked at Microsoft. The culture was different. The way that, you know, management was done was different. The ideas were different. So it's a culture. And I don't know Apple. Well, I know a couple of guys who worked at Apple, but I think that probably Apple and Microsoft are pretty, pretty close, I would say. The thing is, Microsoft is I lived. I actually did the privilege to work under Bill Gates, Steve Ballmer, and Satya Nadella.
00:39:39
So I got the three CEOs, which were three different people. Gates wasn't well, was an engineer. So, he understood what he was talking about very well; product-focused. And I think that he trickled down the whole hierarchy; so it was product, product, product. Then Ballmer took over and Ballmer is a sales guy. And for a couple of years, it was sales, sales, sales. Product was that we sell anything so we could sell brooms. We can sell whatever, sell it. And I was kind of. Very, very it's a it's a kind of shift of a culture at some points. And then Nadella took over and Nadella is an engineer as well. And he brought that product-based thinking and innovation and trying to have a platform play.
00:40:20
So it was not just about bringing features for the sake of features was bringing features in the sake of, well, we're going to we're going to solve problems. Again, I think that's an engineer mindset that Gates had still have and that Nadella had. And I think the culture was really around bringing new things on a constant basis is just innovation, innovation, innovation. Of course, it's still still a very sales-oriented organization. They are making a lot of money there and they are very aggressive from a sales perspective. But the whole culture is really around product, product, product. And that's what I felt pretty well. And that's where you see engineers who are going up the ladder on pretty, pretty high executive positions because they understand that it's a product-driven company.
00:41:03
Oracle on the other side was a completely sales-driven company. They could sell and apologize for the bad words, but it could sell really badly, bad products. They did not care as long as they were selling it, although they are selling good products. But the products were really on the back of their mind. They were pushy, pushy, pushy. So complete two different cultures. And I think that Apple is probably it's a different from Microsoft. And, you know, we had a lot of discussions, you know, that I'm a PC, I'm a Mac kind of thing. And it's true that Microsoft has not been well known for its design, where Apple was, you know, all about the design. But I think from a technological standpoint. I think that both are really focused on creating great products.
00:41:40
But it's a it's a product-focused organization. That's also super interesting that you actually bear witnessing the different leaders, different CEOs that they were changing a lot. Then the navigation of the whole culture of the company. I mean, the product was still the same, but the culture of Microsoft, for example, changed under the leadership of different persons. So. So that's really, really interesting. My last question will be also about your career, because it's a lot about technology. You said that you started with, for example, robotics and then you moved to Microsoft and your career is like super successful. But when you're looking back, like what would you advise to the people who would, for example, want to embark on the similar journey as a.
00:42:36
You are currently in because like there it had to be like a roller coaster, for sure, as for everybody. But like, what would you advise to the people who would like to embark on the similar career journey as you did? Well, I'm not sure I was successful. Success is a subjective matter. I think that I was lucky, first of all. And I think my best advice to young people, and that was the advice I give to my kids as well, is: do something that you like first. If you don't like it, change it. Don't move. Don't work either for yourself or for a company or for for a business that you don't like. So and I think that when I started always and I started as an employee, then creating my first company and I think that I just did what I what I like to do.
00:43:22
So on a daily basis, of course, I learned a lot of things. You know, I was an engineer. We built where we created that company in 1993. Yeah, a long time ago with two friends. I did not know how to run a company. I've never run a company before. And so I had a lot of things to learn. And. And we actually we almost bankrupt the company. We made all the possible mistakes. But, but I was doing something which was just incredible and I loved it. And of course, some mornings there were, you know, bad mornings like everyone else. But I would say that's the first thing. And if you don't like your job, change jobs. OK. And I think that particularly in the I.
00:43:55
T., while in the technology space, everybody's looking for for these days, whether you're in I. T., in A. I., in cybersecurity, you name it, you will find easily job. Or you can create easily businesses, I would say, across the world. But yeah, really enjoy what you're doing. That's the first thing. And the second thing is as well. Second thing is being curious as well. Technology is one of those things where, you know, it moves very fast. As I said, A. I has changed dramatically for the last two years. Every week you've got something new and you probably will find something that interests you more, whether it's agents, whether it's, of course, large language models or diffusion models, whatever. So lots of things there. But be curious.
00:44:35
Be curious is, you know, it's like this Roman Empire thing is just look around and you will find a lot of ideas, a lot of opportunities. Ask Chad GPT if you don't know, give you ideas. But being curious is great. You know, I used to travel a lot. And that's a story I was telling, you know, telling a lot of people. And when I was traveling and I traveled a little bit less now, but I was traveling and I was every time I was flying, I was buying a couple of magazines. And I always try to buy one magazine in a topic that I didn't know anything about. Whether it was, you know, gardening or building houses or whatsoever. And just for the sake of I wanted to understand how other people who are not like me were thinking.
00:45:13
And so it was kind of funny because I just and it was anything. OK. And I discovered that, you know, people were, I don't know, gardening some weird flowers that I didn't know about. So that's interesting. You know, those things. And I think that that makes your mind be very open to potential opportunities. And actually, when I joined Microsoft, I did not intend to join Microsoft. It was just an opportunity that arose. It arose because I met people, said, hey, this is who I am. And then suddenly somebody approaches me and says, by the way, would you be interested in doing that for us? Says, why not? Let's try. And that was not really my field. But I said, OK, let's do it. And then I jumped into the bandwagon.
00:45:50
I started Microsoft Consulting Services in Southern Africa. And that was the beginning of a long story. So being curious is really being open-minded. And the thing is that we need to be cognizant of the fact that we're in 2024. Opportunity. Opportunities are all around us. And so if we are curious and open to those opportunities, they will come. And then you can grab one and run after it. Yeah. And so that would be my best career. It's not, you know, being open, being passionate about what you're doing and the world can be yours. And go to that shop and buy the garden and magazine. Or whatever, golfing or anything. But I really like it because it's also like when I'm flying, for example, people ask me about, oh, what you are reading.
00:46:42
I was, you know, just like looking a bit like at your lines. I really like your book, what you are reading. And then if you are reading the gardening, like you just want to get into this mindset. And then people are like, oh, I also like gardening. Like, what is your favorite? Like, for example, like how you are treating your pumpkins. And I would be like, sorry, I'm just trying to understand. Yeah. But that's. It opens opportunities even like that. So, you know, I don't know a little anything about it. But by the way, you know that. That's cool. Tell me about it. And then you start, you know, having conversations with people that you would never have conversations with otherwise. And I also have to ask, what was your first business about with these two friends?
00:47:23
What was the We were. That was really servicing. Well, IT services. Mostly we were. And think about 1993 is basically the beginning of Internet. It was barely beginning. You know. And I always remember my first email was one four nine zero five two at computer dot com. So it was not even my name. There was no such a thing at that time. So it's just a number. And I got my. So we started really at the beginning of Internet and we, you know, installing servers, mail servers, those kind of things. So it was booming at that time. So it was a we were doubling the revenue almost every month until we spent too much, too much money. Didn't care about cash flow. So we almost bankrupted the company. But yeah.
00:48:05
So there was IT services mostly. Wonderful. Alright. So and that's a wrap-up of today's episode of Create the Future. Now, I hope that you enjoyed the conversation and gained some valuable insights into, for example, Microsoft artificial intelligence. We also dived deeper into the history and how history can actually show us the innovations that we are evolving into. Because from the fire, we got into. The AI and at the end, we even wrapped it up with the gardening. And as always, we are here to help you navigate the evolving digital landscape and create an impactful future with technology. If you love today's episode, don't forget to subscribe. Leave us a review and share it with others who are also excited about innovation. You can also follow Innovatology on social media for more updates, tips, and behind-the-scenes content. Thank you again, Marc Israel. For joining us today and sharing your expertise, and thank you also to our listeners for tuning in until next time, stay curious, as we also mentioned during our episode and also stay innovative. So thank you so much. Thank you, guys. Have fun. Thanks, Mary. Cheers. Bye bye. Bye.
