Artificial Intelligence Research Community and Associations in Poland

In last years Artificial Intelligence presented a tremendous progress by offering a variety of novel methods, tools and their spectacular applications. Besides showing scientific breakthroughs it attracted interest both of the general public and industry. It also opened heated debates on the impact of Artificial Intelligence on changing the economy and society. Having in mind this international landscape, in this short paper we discuss the Polish AI research community, some of its main achievements, opportunities and limitations. We put this discussion in the context of the current developments in the international AI community. Moreover, we refer to activities of Polish scientific associations and their initiative of founding Polish Alliance for the Development of Artificial Intelligence (PP-RAI). Finally two last editions of PP-RAI joint conferences are summarized. 1. Introductory remarks Artificial Intelligence (AI) began as an academic discipline nearly 70 years ago, while during the Dartmouth conference in 1956 the expression Artificial Intelligence was coined as the label for it. Since that time it has been evolving a lot and developing in the cycles of optimism and pessimism [27]. In the first period research in several main subfields were started but the expectations the founders put were not fully real­ ized. Thus, the disappointments and cutting financing in the 1970s led to the first, so called, AI winter. The research was intensified again in 1980s, mainly with promoting practically useful, narrow purpose systems, such as expert systems, based on symbolic approaches and logic [21]. Nevertheless, they were not so successful as it was expected. Then, important changes in AI paradigms concern non-symbolic and more numeri­ cal approaches [1]. During the end of 1980s many researchers focused interests on * Institute o f Applied Computer Science, Jagiellonian University, and AGH University o f Science and Technology, Cracow, gjn@gjn.re ^Institute of Computing Sciences, Poznan University o f Technology, Poznan, jerzy.stefanowski@cs.put.poznan.pl 160 G. J. Nalepa, J. Stefanowski methodological inspirations coming from statistics, numerical methods, optimization, decision analysis and modeling uncertainty. It helped in a significant progress in new machine learning methods, rebirth of neural networks, new developments of natural language processing, image recognition, multi-agent systems, and also robotics [11]. Several researchers proposed new approaches to manage uncertainty and imprecision, while others significantly improved genetic and evolutionary computations which started computational intelligence subfield [10, 7]. All of these efforts led to the new wave of applications, which were far beyond what earlier systems did and additionally boosted the growing interest in AI. Since the beginning of this century one can observe the next renaissance of the neu­ ral networks research, in particular promoting deep learning, and intensive develop­ ment of machine learning together with appearance of Big Data [33]. Other advances were also done in computer vision, improving perception of intelligent agents which can perform more complex tasks. New ways of interactions with human were also developed in fields of Ambient Intelligence and smart devices [26]. Moreover, robotics benefits from the fast pace of advances in machine learning, computational intelli­ gence, uncertainty representation and handling, decision making, and multi agent systems. A strong improvement of perception in robots supported progress in hu­ man robot interfaces, their understanding and learning [30]. Furthermore successful techniques were introduced in speech recognition, natural language processing, au­ tonomous systems and self-driving cars. The trustworthy, human-center AI systems and explainability are of crucial importance in AI based system, as in this area the decisions made by algorithms may have immediate physical consequences, and may put at risk human health or lives, e.g. in autonomous driving. Concluding, the never seen before peak of hype around artificial intelligence has occurred in the last years. However this peak is different than previous ones. It is much stronger and touches different recipients than the research communities only. It seems to look ” like a storm’ changing the world” . One can notice that several factors came together in the last decade: • Several new methods, e.g. deep neural networks, and intensive developments of older approaches led to a scientific breakthrough, • Appearance of Big Data, where large volumes of data, having different represen­ tations, enable several algorithms to be more efficient and surprisingly accurate in solving difficult, real world complex tasks; Big Data is also characterized by other properties such as Velocity, Veracity or other complexities which have opened new research and application perspectives [17], • Increasingly powerful computers with greater storage and parallel processing become available and cheaper; the easier availability of GPU hardware and computations had a big impact on training of deep neural networks, • Advances in solving spectacular real life case studies, e.g. self-driving cars, games such as Go, intelligent query answering and NLP in IBM Watson, medical image recognition, Big Data mining, where intelligent systems could achieve accuracy comparable to humans, Artificial Intelligence Research Community and Associations in Poland 161 • AI techniques were moved from laboratories to industrial practice, which also attracted a wider attention from other communities than academic researchers. Furthermore real financial investments were made by many commercial compa­ nies. It increased the number of real world applications and boosted selling AI-based products, which provided added economical values. Several reports, such as [2, 25], present information showing that the AI sector has become a growing target area for such investments in the last decade. For instance according to [2] private equity investments in AI companies and start-up accelerated from 2016 (e.g. it doubled from 2016 to 2017 reaching 16 USA billion). The reader can also refer to the fifth chapter of [25] for more details on revenues of AI market. These economical aspects constitute a large difference to earlier moments of general interest in AI and its opportunities. Nowadays, many managers, economists, sociologists or administrative officials per­ ceive Artificial Intelligence as a general — purpose technology that will revolution­ ary change the world economy and society. On one side AI applications may improve productivity gain, saving costs and enable better resource allocation. On the other hand, statistical reports of [2] demonstrate that the large scale effects of AI requires investments in a number of complementary inputs (e.g. infrastructure, collected data but also to train a specialized staff). The last year McKinsey Global AI report [18] provides results of a large survey (over 2360 participants from various companies all over the world) showing nearly 25% increase of AI applications in standard business processes, where in over 50% they significantly reduced costs. Moreover, 63% respondents are seeing growing re­ turn from investments (ROI) from the AI adoption. The highest revenue increases are reported most often in marketing and sales while cost decreases most often in manufacturing. This report also shows which AI methods are the most popular in particular domains. Furthermore other pooling results include risk identifications, in particular a limited access to well prepared data, its good quality, along with privacy protection issues. To sum up, nowadays AI is more and more applied in various areas and often produces money returns. One can also informally say that business began to believe in intelligent products. Besides benefits of applying AI, several people (also coming from sociology, ethics, philosophy or law) are considering limitations, risks and ethical issues. While philosophers raise more fundamental questions about what we should do with the fast developing AI systems and robots, what the systems themselves should do, what risks they involve, and how human can control these systems1 or how to relate them to respecting human rights, democratic values. The researchers from other fields consider other risks or limitations such as threat to privacy, security, safeness, legal responsibility2. Changes of human work, replacing or moving people from one to another new job, continuous education and skill development are next elements of societal A I impacts. 1For a brief definition o f research on this field and links to main debates the reader can consult the section entitled Ethics o f Artificial Intelligence and Robotics inside Stanford Encyclopedia of Philosophy h t tp s : / /p la to .s ta n fo r d .e d u /e n t r ie s /e t h ic s -a i / . 2 Many intensive discussions on so called superintelligence and the problem o f human control over so fast developing and more and more powerful AI systems or robots have also been undertaken by researchers coming from various fields for instance see the summary available in [19]. 162 G. J. Nalepa, J. Stefanowski This raises many public considerations about regulations and needs to ensure trustworthy, human-center A I systems. In particular it is visible in European Union experts’ discussions, working polices and several recent recommendations or white papers. For instance last year the High-Level Expert Group on AI presented Ethics Guidelines for Trustworthy Artificial Intelligence. In February 2020 European Commission released a special white paper on AI, which provides their views on the upcoming policy, addresses the risks associated with AI usage, and discusses future regulatory steps on Artificial Intelligence. From research perspectives it opens several new challenges how to incorporate these recommendations into inte