The Future of Work – Artificial Intelligence (AI) impacts on learning, teaching, and education

At the November 2017 Gothenburg Summit, the Commission presented the Communication ‘Strengthening European Identity through Education and Culture’, that set out a vision for a European Education Area and announced a dedicated Digital Education Action Plan, which aims to foster digital skills and competences for all citizens. The Action Plan focuses on implementation and the need to stimulate, support and scale up purposeful use of digital and innovative education practices. It has three priorities: making better use of digital technology for teaching and learning; developing relevant digital competences and skills for the digital transformation; and improving education through better data analysis and foresight. Artificial Intelligence (AI) will have an impact on all these, and in the last priority the Communication specifically invites to explore its impact in education and training through pilots. This policy foresight report suggests that in the next years AI will change learning, teaching, and education. The speed of technological change will be very fast, and it will create high pressure to transform educational practices, institutions, and policies. It is therefore important to understand the potential impact of AI on learning, teaching, and education, as well as on policy development.

AI is currently high on the political agendas around the world. Several EU Member States have declared it as a political priority. Influential studies now suggest that perhaps one in two occupations in the industrialized countries is likely to become automated using already existing AI technologies. Policy makers at the European Parliament have highlighted the importance of the issue, and the European Commission, in its 2018 annual work programme, sets its wish to make the most of AI, which will increasingly play a role in our economies and societies. AI is now often called “the next electricity.” The transformative impact of general purpose technologies, like AI, however, becomes visible only gradually, when societies and economies reinvent themselves as users of new technologies. Technological change brings social and cultural change that is reflected in lifestyles, norms, policies, social institutions, skills, and the content and forms of education.

Wide availability of cheap processing power and vast amounts of data in recent years have enabled impressive breakthroughs in machine learning and created extraordinary commercial and research interest in artificial neural networks, i.e. computational models based on the structure and functions of biological neural networks. Neural AI, and machine learning methods associated with it, are now used for real-time language processing and translation, image analysis, driverless cars and autonomous vehicles, automated customer service, fraud detection, process control, synthetic art, service robots, and in many other applications. Although some of this excitement may be based on unrealistic expectations and limited knowledge of the complexities of the underpinning technologies, it is reasonable to expect that the recent advances in AI and machine learning will have profound impacts on future labour markets, competence requirements, as well as in learning and teaching practices. As educational systems tend to adapt to the requirements of the industrial age, AI could make some functions of education obsolete and emphasize others. It may also enable new ways of teaching and learning.

In the European framework programmes for research and technological development, AI technologies have been studied and applied in educational contexts in many projects focusing on technology-enabled learning. These projects have used technologies that have deep ties with AI research, including natural language processing, pattern recognition, intelligent tutoring, probabilistic AI planning, intelligent agents, AI game engines, and adaptive user models in personalized learning environments (PLE). The impact of these technologies in practical educational settings has been relatively modest until recently. Technical developments over the recent years, however, suggest that the situation may be changing rapidly.

The main intent of the present report is to help educators and policymakers to make sense of these potentially very important technical developments. To understand the impact of AI, we need to understand what AI is and what it can do. In the current “AI avalanche” this is not always easy. Deep expertise in AI technology is scarce, and many educators and policymakers now struggle to get up to date with basic knowledge in this area. In the midst of self-driving cars, speaking robots, and the flood of “AI miracles”, it may be easy to think that AI is rapidly becoming super intelligent, and gain all the good and evil powers awarded to it in popular culture. This, of course, is not the case. The current AI systems are severely limited, and there are technical, social, scientific, and conceptual limits to what they can do. Perhaps surprisingly, well-established research on human learning provides important tools and concepts that help us understand the state-of-the-art and future of AI. Many current AI systems use rather simplified models of learning and biological intelligence, and learning theories thus help us gain better understanding of the capabilities of current AI systems.

There will be great economic incentives to use AI to address problems that are currently perceived as important by educational decision- and policy-makers. This creates policy challenges. For educational technology vendors it is easy to sell products that solve existing problems, but it is very difficult to sell products that require changes in institutions, organizations and current practices. To avoid hard-wiring the past, it would be important to put AI in the context of the future of learning. Policy may be needed to orient development in AI towards socially useful directions that address the challenges, opportunities, and needs of the future. As AI scales up, it can effectively routinize old institutional structures and practices that may not be relevant for the future. Future-oriented work, therefore, is needed to understand the potential impact of AI technologies. How this potential is realized depends on how we understand learning, teaching and education in the emerging knowledge society and how we implement this understanding in practice. Future-oriented policy experimentation, as suggested by the Digital Education Action Plan, may, therefore, be an effective way to address this challenge.

Recent AI breakthroughs are based on supervised machine learning. A critical success factor of these systems is the availability of huge amounts of pre-categorized training data. In contrast to logic- and knowledge-based approaches to AI, we therefore characterize these as “data-based” AI systems in this report. Many of these “deep- learning” neural AI systems may well be characterized as “datavores.” At present, the most important technical bottleneck of AI, therefore, is the availability of data. This is a qualitatively new development in the history of computing and information processing. Without access to vast training datasets, it is very difficult to develop successful AI systems. In this report, we put forward an argument that EU policies could create data platforms that could redefine the competitive landscape for learning- and education-oriented AI systems.

As these supervised AI learning algorithms are based on historical data, they can only see the world as a repetition of the past. This has deep ethical implications. When, for example, students and their achievements are assessed using such AI systems, the assessment is necessarily based on criteria that reflect cultural biases and historically salient measures of success. Supervised learning algorithms create unavoidable biases, and these are currently extensively debated. From a more fundamental ethical point of view, however, the expression of human agency requires capability to make authentic choices that do not only repeat the past. Although there are already AI systems that deal with creative activities, AI systems will have great difficulties in dealing with people who are creative, innovative, and not only average representations of vast collections of historical examples.

It is often assumed that AI systems enable new levels of personalisation and diversity for information systems; much of this, however, results from fine-grained categorization that puts users into pre-defined classes. Although these systems may be able to efficiently simulate personalisation, they do not necessarily support deeper levels of diversity. At present we can say that the use AI systems in educational settings will shape the development of human cognition and self-efficacy, but we don’t know how. It is therefore important to continuously evaluate, for example, how the use of AI in educational contexts constrains and enables human possibilities for responsible and ethical action. AI systems can be excellent predictive machines, but this strength may be an important weakness in domains where learning and development are important. A contribution of this report is to show that different types of AI and machine learning systems operate on different layers of human behaviour4. Most importantly, the level of meaningful activity—which in socio-cultural theories of learning underpins advanced forms of human intelligence and learning—remains beyond the current state of the AI art.

One of the most successful application areas in AI has been video processing. There will be strong economic interests in using video-connected AI systems in classrooms and to complement the collected data with data from social media and Internet of things (IoT) platforms. As it becomes technically possible to monitor student emotions and attention in real time and use such data to help teachers and students, AI privacy and security become important topics also in education. Similarly, AI systems are well suited for collecting informal evidence of skills, experience, and competence from open data sources, including social media, learner portfolios, and open badges. This creates both ethical and regulatory challenges.

Several high-profile econometric studies on the future of work have shown that many occupations can be automated with current AI technologies. These studies have relied on task- and skill-biased models of technical change. In this report, we argue that a data- biased model is more appropriate for current AI systems. We also explore a similar methodology to see how the future of the teaching profession might look like. The results suggest that many currently defined high-priority teacher tasks might be automated. However, this is based on the assumption that the role of teachers is rather mechanical and purely instructional with summative assessment playing a central role, reflecting deep beliefs about the functions of education and the social institutions around it. In educational systems that emphasize development and, for example, social competences, formative assessment might be higher on the list. As a result, there is a risk that AI might be used to scale up bad pedagogical practices. If AI is the new electricity, it will have a broad impact in society, economy, and education, but it needs to be treated with care.

Chosen excerpts by Job Market Monitor. Read the whole story at The Impact of Artificial Intelligence on Learning, Teaching, and Education | EU Science Hub

Related Posts