Pretplatite se na newsletter portala Euractiv.hr i prvi saznajte novosti iz Europske unije.
1. How do you describe the environment for the development of artificial intelligence in Croatia?
About three years ago, thanks to a similar question such as this, we concluded that it might be best to start gathering as a community around AI development. We started with a small, interesting group of enthusiasts from startups, big tech companies and academia. Today, we have a community in Croatia that brings together around 4,500 people, and together we push the development and application of artificial intelligence through technological solutions because people are interested in. The potential that AI has and the economic value that can be generated through AI is quite large. Then we raised additional questions - legal, moral and ethical.
The position of AI in Croatia is getting better - primarily because of people who are building solutions on AI platforms. I would say that the younger and modern generation stopped expecting what my generation may have expected, which is: what is the role of the state, what is the role of the regulators, what we can expect from the European Union etc. They don't want to wait and stand. These are young companies and people who are globally competitive and have skills that can be applied globally. They have their own path and a successful international path, and they would like to connect with as many people in Croatia as possible - not only in the field of development but also in the domain of AI.
2. What is currently being done in Croatia regarding the development of AI? Why do we get impression that AI does not even exist in Croatia, that entrepreneurs work on this issue while the government is slow in implementing strategies?
Croatia has a strong attachment to AI development within EU plans and activities. The EU is unfortunately too focused on regulation, setting up an environment, addressing how AI can be programmatically pushed forward - which is not necessarily the fastest way. In April last year, a Declaration on cooperation in AI was adopted, which Croatia subsequently signed after several months. On this basis, we, as a state, are obliged to bring at least two documents: the 'Strategy for the Development of Artificial Intelligence ' and then its 'Application in the economy' , which is what the Government and the Ministry of Economy is currently working on. These are normative things which will be necessary in applying for EU funds, where we will then be able to apply for certain pillars of investment with our companies and our solutions. One of them is the pillar of artificial intelligence. In this pillar, € 2.3 billion is planned for AI investments within Digital Europe 2021-27 .
The state has a role to play, but unlike some other industries, we are not necessarily dependent on it, but at some point it will be a good generator of development. I am sorry it is perceive that there is no AI in Croatia, and that we are not dealing with it. The reason for this, I would say, is that our companies that are working on AI are more oriented on international export as it is the case with all IT companies. They are doing very interesting projects in Croatia and abroad in various domains of industry, primarily in health and transport. It is the question of time when the Croatian economy will be more open to technology, so we will see more use of AI in our country - it depends on several factors.
Srdjan Vrancic / CROPIX
Whether artificial intelligence and technology in general will make some jobs disappear, answer is 'yes'. Will it be massive? The answer is 'no'.
3. Would you say that Croatia is not currently making enough use of EU initiatives to develop AI, such as Horizon?
I think this is a general question of how much Croatia is using EU funds at all, and consequently even some specific ones. This year, we started on time to prepare for the next financial period, while we joined the previous one at the very end. The goal of IT professionals is to bring closer the platform and capabilities through the Programming document that we have brought in HUP and through conferences dedicated to AI. This way we will be ready when opportunities arise and we will have project completed within the framework of which the state will enable us to access those funds. I think we are in a better position today than we were with the first funds when we joined the EU.
4. What about our strategic plan for AI, since Croatia is among a small number of Member States that have not yet submitted their plan to the European Commission? What do you think the AI strategy in Croatia should include?
I think we must first demystify the term Strategy. It should be a several page document that generally directs 4-5 questions that are of interest to us in the field of AI and in some program plans involve deeper activities. We are used to call a Strategy document which has 400 pages, and which is all but not a strategy. The positive thing is that the Ministry is working on producing and releasing this document. Of course, the best approach to drafting documents of this type is to properly involve the entire interested community, right from the very beginning of document creation. Currently it is a closed document within the National Digital Economy Council's Working Group . We expect - when I say 'we' I mean employers, the community and the academic sector - that this month we will get involved in commenting and collaborating further on building this strategy, as well as other documents needed to better deploy AI. Unfortunately, I consider this to be an old and traditional approach to creating such a document, I think we should work on that document right from the beginning. Other countries are already working on the next version of the strategy and even the Action Plan, while we have not yet adopted the first one.
Srdjan Vrancic / CROPIX
Declaration on cooperation in AI obliges that we as a state have to bring at least two documents: the 'Strategy for the Development of Artificial Intelligence' and then its 'Application in the economy',
5.What aspects need further work to get closer to some of the current AI leadership members. What did France, Germany, the United Kingdom or Estonia do differently to become some of the frontrunners?
They already have the foundation for it. It's easy to say 'we would like to be the frontrunners', but in our domain the AI domain is just emerging. France, United Kingdom, Germany come from the field of digital transformation, digital economy, industry 4.0. They have already spent ten years in the mental frame of change and transformation, and for the last ten years we have been thinking about how to set ourselves up for it.
France opted for AI as one of the fundamental things and has put all other strategies, like industrial strategy, according to AI. In Croatia, we still need to revise these strategies to bring them into line with these new trends and the technologies that are coming to them.
6. I have heard among some entrepreneurs that the EU is raising barriers and limiting innovation with regulations, privacy or copyright laws. Does this pose a problem to European companies globally as opposed to the situation in China or the US?
There are several different approaches in the world related to the building of capabilities and the application of artificial intelligence. They are led by China, the USA and now the EU is trying to join it. When looking at AI-relevant data, in China, everything related to cognitive services does not raise ethics, morals or rights issues . If something has a state goal then it has a state goal. USA is very practical. They approach with the principle: 'let's do it first , and after we will regulate it'. This is why some technologies are now revaluing and raising questions about privacy and security. The EU has traditionally been very standardized. EU will spend the first few years discussing the legal framework and the laws that need to be put in place for someone to do something. This is a drawback of the EU - we are over-regulated and over-normalized . Probably the companies that ask these questions are right to ask them. They come from the EU domain and say that we are losing this race because first we have to regulate the frameworks in which the construction and application of artificial intelligence takes place. I would say, from the domain of privacy protection, this certainly has its good backing. But from the domain of development, we need to find our boundaries as soon as possible in order for it to function more simply.
7. Can you list examples of how AI can contribute to Croatian industry, how it can boost economic production and development?
In the context of the document framework on 'AI Potential in Croatia' that we have brought within HUP, the next step we need to do with individual industries is to answer the question: What would it mean to apply AI in a specific industry? This is the next document that we are working on and that will provide concrete indicators of the expected economic contribution of AI. What I can say for sure is that we already have some pilot projects in Croatia deploying AI in domains that are sometimes aligned even with smart specializations. We have several companies working on a combination of AI in terms of transportation . We have companies that work on so-called RPA models (Robot process automation). We have successful companies working in the healthcare field where AI helps specialists and doctors point out specific anomalies that they need to notice and that they need to recognize. This is easy for AI because it can more quickly identify a potential patient problem based on historical data.
Srdjan Vrancic / CROPIX
Ratko Mutavdžić speaking during the conference AI2FUTURE in Zagreb, Croatia
Then where is this boundary between AI and the individual?
The fact that the doctor in healthcare is the decision maker is working relatively well so far. I do not see that AI will soon be left to make some important decision in healthcare, but it will help the doctor to be more efficient and faster. They will prepare the data area for him to make a better decision. Furthermore, in the area of justice in countries such as Estonia, small fines are automated. The Ministry of Justice has decided that because of small fines, such as a speeding fine of HRK 500, you do not have to go to court but the offense can be done automatically on the basis of pre-existing cases and punishments that automate the process.
AI is not just a matter of big processes and big cases, but also of the daily use of devices and programs that have elements of AI in them. If you take a photo with your cell phone and wondering how good out, iti s because of the AI and not because your good eye or the professional quality of the photographer. We have accepted this chip placed in a cell phone that produces a better image for us. I see more and more examples of where AI will start to apply around us, and then at some point we will say: 'this is really the AI behind it'.
8. The unemployment rate is really high in Croatia and often it is said that with the development of AI many jobs will be lost. How much truth is there and how much misconception in this fact?
As you said, there are so many unemployed and unaffected by modern technologies - but modern technologies and artificial intelligence will transform jobs. People will stop doing repetitive jobs that will be automated, and they will do something better and more meaningful than that. Generally, we make such changes with digitalization and digital transformation, and already today you expect that no one has to issue you a certificate of Croatian citizenship, but that either the state will do it for itself or you will create a digital certificate through an electronic portal. Whether artificial intelligence and technology in general will make some jobs disappear, answer is 'yes'. Will it be massive? The answer is 'no'. So what is going to happen is that jobs will have to be transformed, and we expect that from digital transformation which some countries have already done for the past 10 years. When you look at some reports the average number of industrial robots per ten thousand employees, we are at the beginning and we have 4, while Germany has 400. So Germany has a hundred times more robots and has lower unemployment than us. It is hard to say that technology will ruin us, it is more realistic to expect that technology will be a platform for change that is necessary and we must make it anyway.
