Uwaga, odcinek nagrany w języku angielskim.
W najnowszym odcinku podcastu 99 Twarzy AI miałem przyjemność rozmawiać z jednym z czołowych ekspertów w dziedzinie sztucznej inteligencji – Stefano Rebattonim, dyrektorem generalnym IBM na region Europy Północnej, Środkowej i Wschodniej. W trakcie rozmowy poruszyliśmy kluczowe tematy związane z rozwojem AI, przyszłością tej technologii oraz odpowiedzialnym podejściem do jej wdrażania.
Najważniejsze tematy naszej rozmowy?
- Etyczna sztuczna inteligencja: Stefano podkreślał, jak ważne jest, aby AI działała w sposób przejrzysty i odpowiedzialny, zgodnie z zasadami etyki, aby eliminować uprzedzenia w systemach AI i wspierać interes społeczny.
- Wyzwania edukacyjne: Rozmawialiśmy o konieczności zwiększenia świadomości i kompetencji zarówno twórców, jak i użytkowników AI, aby technologia mogła być właściwie wdrażana i rozwijana.
- Rola IBM w sektorze B2B: Stefano wyjaśnił, jak IBM koncentruje się na dostarczaniu rozwiązań AI dla biznesu, które pomagają poprawić efektywność i produktywność, zwłaszcza w sektorach takich jak finanse i administracja publiczna.
- Regulacje a innowacje: Omówiliśmy również wpływ regulacji, takich jak unijne prawo dotyczące AI, oraz jak znajdują one równowagę między innowacją a bezpieczeństwem.
- Przyszłość AI: Stefano podzielił się swoją wizją przyszłości, w której AI i technologie takie jak komputery kwantowe mogą zrewolucjonizować różne sektory, w tym bezpieczeństwo i badania.
Dobrego dnia! Posłuchaj podcastu i podziel się nim z innymi!
Tradycyjnie wielka prośba – jeżeli jeszcze nie wystawiłaś lub nie wystawiłeś oceny 99 Twarzom AI na Spotify lub Apple podcast – pomóż mi proszę z dotarciem do nowych słuchaczy. To zabierze Ci tylko 2 minuty. Z góry serdeczne dzięki!
Dobrego dnia i niech #AI będzie z Wami wszystkimi!
Transkrypcja rozmowy
Disclaimer. Droga Czytelniczko, Drogi Czytelniku – mała uwaga dotycząca transkrypcji rozmowy. Jak się pewnie domyślasz – transkrypcja została przygotowana z wykorzystaniem magii LLM. Proszę o wyrozumiałość, gdyby pojawiły się w niej niewielkie błędy, literówki, etc. Żeby mieć pewność co do wszystkich wypowiedzi – polecam posłuchać, zamiast czytać. Ukłony.
Karol: Stefano, it’s great to meet you. It’s great to have you as one of the 99 faces of AI, the podcast I’m producing. Thanks for having me. Thanks for the invitation. For those people who don’t know you, who are you. What’s your AI story, Stefano.
Stefano: That’s a good question. So first, I’m an IBMer. I’m a second generation IBMer. My father worked for the company in the 90s. I’m just been appointed as new general manager for the Northern, Central and Eastern Europe for the company. Congrats. Before that, thank you very much. As probably my accent is saying, I was in Italy and I was running the business for IBM Italy in the last four years.
Karol: I will repeat my question. What’s your AI story?
Stefano: Oh, it’s a story of, you know, passion. It’s a story of being very much involved in conversation that are spanning through institutional levels from ministers, ministries. to private company to public administration on really to try to put ai in practice in the sense on which IBM is looking at that so a sense of ai for business so the application implementation of artificial intelligence not per se but for something that is bringing augmented productivity augmented efficiency more quality of services and products So how to educate the market.
That’s the question, honestly. I’m used to say that technology per se is not enough. So technology and human capital are two legs that have to move at the same rate and pace. And one of the most important thing is really to elevate the understanding, the knowledge and the competencies of not just the developers, but also of the users. Because most of them, if not all of us, should be required to interface and to use these algorithms.
Karol: Without awareness and without knowledge of what is possible, there is no demand.
Stefano: Exactly. And that is also a pretty risky part we need to be careful about. Because from one side, the user has really to enable their competence to better understand this technology. But from the other side, the developers, so who is working on that from the offering side, should really have a lot of attention in things like ethics, transparency, and demonstrability of these algorithms. Because we do not want to have a technology that is taking decisions without being able to demonstrate why and how they get to the decisions.
Karol: I will share a kind of insight with you. Do you know which of the episodes, how do you think, which of the episodes are the most popular ones in my podcast. The one around ethics, I think. Yes, exactly.
Stefano: Ethics and the basics. Yeah. So let me just say that on this, IBM has a long history of, I would say, reputation that has been built around that. We were among the very first partner in the so-called Rome call for AI ethics. And I was pleased to be in the Vatican in front of the Pope. Yeah, it was a big thing. It was a big thing with our chief research officer, Mr. Dario Gil, signing the commitment of IBM for developing an ethic and responsible AI. And that was made publicly because we really wanted to show up as being a responsible provider in that space.
Karol: What does it mean to be responsible in building AI solutions.
Stefano: So we are referring here to disruptive technologies.
As always, I think that the important thing is to not leave anyone behind. I think that the implementation of AI should be pursued in the interest of the collectivity. So having not a technology that is conditioned by bias or is basically getting decisions that can even more divide within very rich people, within very poor people, but really to try to elevate the level of the society, the level of the well-being for everyone. And I think that this is the concept that is behind our commitment towards the Vatican Academy. So the thing that we want to be transparent in the algorithm that we design and we define, we want our algorithms to be demonstrable. So which are the data on which they work on and why did they get to that decision. And we want to be very much bias-free. So there is a science behind that that demonstrates that consciously or unconsciously, people, when they do have to develop anything, they are putting some bias that they are getting from their personal experience. And we get serious to this point up to indemnifying our clients and partners if there is anything that goes wrong with our larger language model.
Karol: My feeling about IBM is, tell me if I’m right, that you’re the company focused on B2B solutions. How does it make a difference between B2B approach and B2C approach, speaking to the confidentiality, privacy, data quality, etc. .
Stefano: It’s a big difference to me. You correctly pointed out, and I told you before, AI for business is very much related to what we are used to do. So B2B, basically. We acknowledge that there are different tools for AI. So tools that are general purpose or tools that can be very core and specific.
We are in the second part of the game. We try to help our clients and partners how to infuse AI in their current IT environment so to evolve them to be much more productive, much more effective in the way they do their business or they operate their services. Most of the time these are refers to very critical services. Let me provide you some examples. We are referring to government services, to public services, to financial services. And so the liabilities that you are required to give and to your customer, to your partners are much, much higher than the one for the final end user.
Karol: And to restore space for a hallucination.
Stefano: Yeah, and there is that risk of hallucination. And then there comes requirements because, of course, everything that comes from the UAI Act. So why are we so serious in doing that. Because we developed our large language model that are strictly aligned with the requirements of the UAI Act so that anyone that could get to our solution in that space, what’s the next AI, what’s the next data, what’s the next governance, can really be safe. in the way they operate with these tools.
Karol: Safety is one side. And creativity.
Stefano: Oh, that’s a very good point.
Karol: My question is, when we look at the regulations in the European Union, speaking to the AI Act, and comparing that with the legal environment of, let’s say, China and the US, Aren’t we taking the right direction to regulate the market too much.
Stefano: Yeah, that’s something that was summarized in a sentence that was, US innovates, China replicates, and Europe regulates.
I like that. But I think that the work that has been done in the EU, to which IBM was part of, was bringing up to a good point of balance. So this concept of risk-based approach or precision regulation in the application of the AI tools and the AI algorithm is what is really trying to preserve the creativity. So there will be some application of AI for very low-risk services that would not be even regulated. And there, I mean, creativity cannot have any border. But of course, when you come to that very specific and critical domain, defense, for naming a few, you have to be very, very serious in how you want to fence that part of the implementation of this technology. And I think that this is something that when it will come at the country level, here in Poland or everywhere in the EU, will even be more applicable to the specificities of the country.
Karol: You know, I was really surprised. A couple of months ago, I was organizing the kind of inspirational workshop for the IT team for one international company. They gathered the IT team from all the countries from the region. And my surprise was that people who run the business, who use SAP, every solution possible, are not familiar with AI.
Stefano: Really. Yeah.
Karol: Wow. I feel that there’s a huge gap and we are kind of living in the AI bubble. The people like you and me who are really following the news developments, the products, etc., Don’t you have a kind of feeling that the IT sector, which is not connected with AI directly, stands behind.
Stefano: So I feel at the moment of truth is pretty close because there was this AI hype that was driven by mostly the implementation of AI. November 22. Yeah, exactly. That was something that technology that was easily accessible to everyone. And this was creating a lot of momentum in the market. So most of the clients and the partners were engaging us just to say we have to do something on that. Okay. And that was, you know, the playground that was basically 2023. So a lot of pilot, a lot of experimentation. Everybody was asking for that. Not just IT. Let’s do everything that’s possible. Yeah, line of business.
I mean, coming from everywhere. The point here is really to define the strategy of adopting AI. And the question is, how many organizations today have clearly in mind which is their strategy in adopting AI. That is not just AI per se. But is AI that is going to be utilized internally or in their external supply chain or how to engage differently with their customer base. And to me, this is something that is requiring a lot of efforts and work jointly not just on the IT function, but overall in the company. And this is pretty challenging.
Yep. And data. Exactly. And understanding which are the data available, which is the quality of the data there are inside the company. which are the services you really want to evolve, which are the ones that you want to get rid of. So it’s a way of re-engineering, not just innovating or modernizing or optimizing per se. And this is the challenge. I think that here, the company that will be more equipped to get there, to define an AI strategy of adoption and define a clear roadmap to get there will be the one that will get competitive advantage in the years.
Karol: Do you have any specific cases that you can share with us from the IBM perspective that are really close to your heart and that shows the possibilities and the power of AI.
Stefano: There are a couple of them. First, there is this concept of client zero. So it means that before getting out in the market, we use by ourselves our AI tools for everything that relates to internal IBM services. And we have been applying AI to most of our HR, operation, finance department and organization.
That was an important initiative that was laid out and developed by straightly from the corporation. That get us to be much, much more productive, much, much more efficient. And so to freeing up people from non-productive, non-differentiated activities, just thanks to what what’s the next was doing for them. And so really to elevate and to shift their competencies to much more highly valuable spaces. And this has been done.
The results are terrific. 20% of productivity gained, hundreds of hours of people that have been pre-upped for valuable activities. And this is moving on. And as far as we got to these initiatives, we go outside in the market, we pitch what we have been accomplishing with our HR leaders, with our operation leader, just to see if there is a way to get there also for our customer and partners.
Then I know that here in Poland you are pretty passionate about sports. So I think that a couple of things that could be very practical in terms of understanding what can be the implementation of AI have been based.
Karol: Speaking to sports, sorry to interrupt you. Do you know the company that is called Respo Vision. Nope. You should definitely look what they are developing right now. They’re working with the international sports leagues, with media companies, speaking to the, analyzing the sports data, speaking to the football matches, soccer matches, as well as creating the vision of the match.
Stefano: Okay.
Karol: Look, when you are producing the content from the football match, you get six cameras, right. They are able to create the view from every point of the stadium in every second. Definitely worth seeing.
Stefano: Yeah, yeah. And this is very aligned with what we did, what we are already doing for the Wimbledon tournament. Probably this is not something that everybody knows, but we are partnering with Wimbledon since more than 30 years. And all during the years, not just the weeks of the tournament, there are people of IBM that are developing the new fans experience to be ready to provide customized analysis inside statistics forecast of their favorite player. But the most important thing was done this year was really to get a kind of autonomous editing of all the matches that were played and in the very first day of the tournament there are more than 16 fields and you cannot have commentator around or having people that are just getting the best image and so having the possibility to hear the crowd reaction and getting into AI was basically at the end of the day getting the best of the day that was edited autonomously by AI and And this was pretty powerful also because then was provided through up to all the fans. Other sports, that is soccer, that is football as well, we made a pretty good couple of partnership with a couple of teams, one in Spain, Sevilla Football Club, and one in Italy, Empoli Football Club. to apply AI in terms of their talent scout research, so to be able to identify which are the players, especially in the very early ages of their career, that could fit best with the game of that team. Are there any success stories. We just started, but this is something that both teams are looking very much with interest.
Karol: Just a second, guys. If you right now would be looking at the eyes of Stefano, they are just shining.
Stefano: Yeah, you know, because football is a very competitive sport. But of course, not all the teams are equipped in the same way. And the beauty of that is that if you can really rely on data that are qualified and powerful tools, we can try to bridge this gap, just working better on which is the identification of the talents before the others to get there.
Karol: Okay, so we’ve got Wimbledon, we’ve got the football clubs. How about US Open.
Stefano: As well, of course, everything that has been done with Wimbledon in terms of client experience will be replicated soon in the US Open that is about to start. And I know that there is Iga Sviotek, that is one of the best players. Our national treasure. Yeah, yeah, yeah. And we have Yannick Sinner as well from Italian tennis player. So let’s see how it will end up. And Augusta Master, so golf. So everything that is related to, we used to say… I’ve heard that.
Karol: Could you tell me more about the case with the golf.
Stefano: Yeah, yeah, because at the end of the day, everything is around data. In fact, we used to say, you see sport, we see data. And the point is how really to rely on this data to provide to the fan, the supporter, those level of information they are looking for in terms of having much more insight, much more statistics, much more forecast, the likelihood to win of each player based on not just on past statistics, but on the way they performed during the tournament. And this is pretty exciting because, you know, you are, at the end of the day, having a different experience with the sport you are passionate about. You get much more info, you understand better the game, and this is something that I think will be the next generation of sport and will become the fun experience for most of the sports.
Karol: Kind of philosophical question, Stefano. How do you think, how did technology will change the business are growing. Will the market become more monopolized or we will have more examples of solo partnerships and single initiatives.
Stefano: That’s a very good question. I think that given the fact that we today operate in a digital economy, everything is digital. So everything is digital, every services, every critical services we are using. I think that the only way to really succeed in a market, in a country, at global level, is really to promote an action that is being participated by all the major stakeholders. And I try to be very practical in saying that. What do you mean by that. Private-public partnership.
Oh, okay. So I truly believe that as for the capacity that digital can have in order to transform to the services, the quality of products, the level of life of the person. It is definitely needed that government, institution, public administrations, companies such as IBM plus its partner do really contribute to a national strategy to really bring this technology to a level that is value for everybody. I know that here in Poland, as it was in Italy, there is a national plan of recovery and resilience in place. And I tend to think that a huge part of that would be related to digital innovation and transformation of the country. So I think that this is a very important topic that needs to involve around the table all the best practices. And what I want to tell you is that these are the things that IBM is best of doing. So really trying to understand how to connect the dots and now defining these enabling technologies really to bring common value to all the society and the country as well.
Karol: Can you give some examples of how you work locally in Italy between IBM and, for example, Ministry of Digital Affairs. Yep.
Stefano: What are those best practices. Of course, we have a government regulation affair organization that was deeply used in Italy in order to connect ourselves with relevant stakeholders. We are doing the same here in Poland.
So in my two days here in Poland, we are meeting several ministers just to try to show up in the same way. We were trying to bring our point of view on things that, based on our experience, not just in Italy, but, I mean, being a part of a multinational company, we saw in strategic areas such as welfare, such as fiscal, such as tax, such as transportation, And we tried to propose initiatives that were not just IBM made, but IBM plus other partners. And this was something that at the end was resonating and was creating the demand for new service. Then, of course, there were RFP, there were tenders, there were all the public purchase and procurement processes to go through. But I think that getting all the relevant stakeholders around the table to think and work together on these specific topics is really the way to not leave anyone behind and promote competitiveness for the country in the market.
Karol: What will happen in two years?
Stefano: I think that… In three years, what’s your vision. My vision is, first, AI is going to become a reality. So otherwise, I mean, we will talk about AI in two years from now, such as the metaverse. As you know, I mean, two years ago, there was this huge bubble around metaverse, and now nobody speaks anymore about metaverse. I don’t think that is the case of AI. First, because… I’ve solved my meta quest.
Karol: Yeah.
Stefano: that’s the proof that’s the proof of that um so sorry ai will become a reality and secondly looking from the eye of IBM i think that we will be very close to another important exponential technology to become reality this quantum and to me we had a conversation with uh from the krakow office from IBM krakow office about the quantum computing as well That’s another big thing. That’s another big thing, especially when you are referring to initiative at country level, because looking at the cybersecurity space, looking at the research space, that’s a revolutionary technology that is not going to substitute all the others, but they are going to work in conjunction with the others. And I think that there I saw a lot of investment coming in. I’m wondering if there will be anything specific also here in the region. Looking forward to that.
Karol: Looking forward to see you more often in Poland. Good luck with all your project ventures. And Stefano, may the AI be with you.
Stefano: Thank you very much. And thanks for having me again. Thank you very much.