Horizon Europe (2021-2027) is an ambitious EU research and innovation program. Particular attention has been paid to groundbreaking research, including the AI Centers of Excellence (AI CoE). This research requires the joint work of specialists in various fields: artificial intelligence (AI), big data (BD), machine learning (ML), etc. Collaboration here is not only the set of work done by the participants of the project, either understanding the logic, methods, and ways of doing the work by other members of this project. Consequently, AI CoE also teach research group participants and contractors everything needed for an in-depth understanding of their project. In order to do excellence job, you need to have excellence trained staff.
The aim of this article is to describe briefly the functions of AI CoE, show its topically and the necessity of its research today. Quite a lot of literature was used to write this text, which was not cited as it would distract from the merits of the matter.
Center of Excellence CoE
The term CoE is as in vogue today as AI. Even regular studies, e.g., of languages or food preparation, have come to be called CoE if they are conducted online without a specific timetable. Due to its wide application and unclear legal precedent, a CoE may have very different characteristics in one context than in another. We are interested in problems in areas such as technology (e.g. Java), business concept (e.g. BPM), skill (e.g. negotiation), science (e.g. health), etc. In academic institutions, CoE is often a team that focuses on a specific area of research, often involving researchers from other fields. CoE can also be inside the organization as a center of competence or capability. Then it consists of employees from a department or a common facility. Such a CoE may have a coordinating function, ensuring that change initiatives are implemented through standard processes and competent personnel. In CoE technology companies, the concept is often associated with new software tools, technologies, business concepts, etc.
Why AI CoE?
The idea for AI Centers of Excellence arose after many years of trying to improve the performance of companies. Even highly paid data analysts hardly influenced it due to the lack of supporting infrastructure. In other words, there is a gap between strategic planning and the decisions made by executive management and the teams that will implement those decisions. A global survey of over 3,000 managers, as well as interviews with executives and academics, showed that organizations see a solution to the problem in the use of ML by their employees.
Only 10% of businesses today derive significant financial benefits from AI technology, in part because the practice of incorporating AI through AI CoE into an optimal system management process is very laborious and demanding. The strategic goal of AI CoE is to get to know the organization, not just to assess its condition. This allows you to optimize the process of adapting to changing conditions, consciously changing processes, broadly and deeply, to facilitate the organization's learning by AI / ML. This learning is demanding: not only human and machine collaboration is needed, but also mutual learning in different situations and contexts. Then the cooperation between people and machines is effective: explore massive amounts of data, look for anomalies and trends in your data, and then present your results in context. Almost seamless AI / ML should be the ultimate goal of all platforms. The AI skills gap will persist and organizations will reflect on new ways to adapt. Widespread use of these solutions will take place once confidence in the underlying technology is built and the end user understands the factors influencing the projected state of the system.
4 AI CoE centers of excellence are currently developing in Europe: AI4Media (AI for media service), ELISE (research network for science and industry), Human E-AI-Net (supports human-level interaction technologies), TAILOR (academic-public- industrial research network). It is planned to organize the next AI CoE on the basis of which the perfect network will be created. In 2021, AI will be gradually and permanently introduced into larger areas of our work and private life, so reducing the risk of companies will become one of the top priorities. The use of AI also brings new serious problems that increase the company's reputational, regulatory and legal risk. AI companies can bypass human rights, for example: discrimination, surveillance, lack of privacy and security, algorithm bias, including demographic: race, age, gender.
Indicators of excellence
Peter Drucker (1909–2005), a forerunner of modern management, said, "If you can't measure it, you can't improve it." At the beginning we are dealing with a far from excellence system. The process of gradual improvement of the system is carried out by a well-organized group of specialists. The system's proximity to excellent is assessed using the key performance indicator (KPI). In recent years, many centers have dealt with the problem of defining such indicators, usually unique for different systems. The problem arises from the necessity to jointly analyze large amounts of data from various sources in order to further develop the system (enterprise, physical facility, society, environment, etc.). It is especially important to achieve a balance between KPI tactical and strategic, operational and financial as well as forecasted.
Management and the leadership problem
Market volatility, digital transformation and innovation are changing the way companies compete in every industry. They increase the demand for business leaders who are the creators of global changes. enabling the creation and renewal of the competitive advantage of their organization.
The problem of higher-level managers is long known. The number of potential leaders in the "key leader age" of 35 to 44 will decline by 15% over the next decade. Business decision makers need to rethink their leadership development strategies. The role of the leader is huge. Data literacy enables a quick path to analyze and indicators that are collected and interpreted in seconds, not weeks. For example, in business, thanks to this, they can operate in accordance with the current needs of customers, suppliers and the market. This starts with the realization that such speed and vision are necessary, and it becomes the leader's responsibility not only to understand how to define measures of excellence, but also to ensure their adoption and implementation. The set of KPI does not have to be large, but they do need to be used effectively. Such an analysis of KPI data is now possible with the help of ML.
The leader's task is also to create a high team culture. The culture in the company is a common set of values (which we care about), beliefs (which we believe to be true) and norms of behavior (how we make things). Culture exists to coordinate efforts, define a common goal, increase predictability, and teach the group to correctly judge what works and what doesn't. The lack of an organized culture can lead to the fact that employees' thinking and behavior may go in the wrong direction and pace. The leader must take a proactive approach to this as early as possible in order to build the right culture and avoid having to transform it in parallel with the transformation of the entire organization.
This article only covers some of the AI CoE concepts and issues. It is a field that is still little known and promises to open the secrets of nature using the combination of AI / ML to process the huge amount of data around us.