KANBrief 2/22
The German government is seeking to exploit AI „for the benefit of all“. Occupational safety and health is one of the areas impacted by this technology.
Artificial intelligence (AI) is viewed as a key technology of the digital age. As yet however, experts have failed to agree upon a universally valid definition of AI. Some consider it a sub-sector of information technology dealing with automated analysis of large volumes of data (big data), machine learning and, by extension, the „intelligent“ behaviour of machines such as robots. Others also group systems that are less complex or not data-driven under the heading of AI. Whatever the definition, the expectations are high: AI systems such as self-driving cars are expected to have perceptive sensors and communicate with their surroundings, and take decisions for human beings.
Many experts consider the possible applications of AI to be virtually unlimited. As a result, the technology has been hyped in recent years, even though it has been under development for decades and has already experienced lean spells („AI winters“) characterized by sharp declines in research funding. Achim Berg, president of Bitkom, the German IT industry association, nevertheless considers AI to be a turning-point for industry and society as a whole. His view is that „we must harness the wide range of opportunities offered by AI for all areas of life – whether in medicine, mobility or education“, since practical application of the algorithm-driven technology is lagging.
Even so: governments are treating the technology as being of crucial importance, and a race for global leadership has begun. Since publication of the US government‘s AI strategy in 2016, if not before, other countries have followed suit in seeking ways not only of promoting research and development, but also of commercializing the technology and catching up with the „AI world powers“, namely the USA and China.
The German government published a paper describing the key points of its AI Strategy in July 2018. In this paper, it noted that „over the past few years, artificial intelligence has matured considerably and is becoming the driver of digitalisation and autonomous systems in all areas of life.“
The German cabinet adopted its national Artificial Intelligence Strategy at the end of 2018 and updated it comprehensively in 2020. The strategy describes various ways by which „AI made in Germany“ can be made world class and established as a globally recognized quality mark. As stated in the strategy, the issue is that of individuals‘ freedom, autonomy, rights to control their own data and make their own decisions; and about hopes, fears, opportunities and expectations. At the same time, however, it is about new markets for German companies, global competition, and Germany‘s future as an industrial location.
The German government is thus seeking to facilitate „responsible development and use of AI which serves the good of society“. Three billion euros is to be made available for this purpose up until the end of 2025. The plan foresees swift and sustainable investment, both public and private, in applications. Whether in medical and environmental technology, manufacturing, service industries, at the workplace or in the form of smartphone apps, AI has the potential to make everyone‘s lives more innovative, smarter and more personalized.
The German government has set itself the goal of making „comprehensive use of the innovations triggered by the technology for the benefit of society at large“. Increasing interaction between human beings and machines/computers requires the highest safety standards to be observed, and hardware and software manufacturers to incorporate these standards directly into their products. The government regards confidence in the technology as being dependent upon AI being explainable and transparent.
In the future, AI will play a significant role in more and more areas of our lives. That is the conclusion drawn in 2020 by a study commission convened by the German parliament (in German). AI functions stated as examples in the commission‘s final report include recognizing speech commands, filtering out spam e-mails, recognizing images, sorting search results, correcting spelling errors and suggesting products. AI systems translate texts and play Go or chess, and they have been beating human beings at the latter for some time now. The technology is used to control vacuum cleaner robots, advanced driver assistance systems and entire manufacturing plants. It increasingly helps doctors to diagnose medical conditions and select the best treatment for the individual patient.
One chapter of the report is dedicated to the world of work. The report states that AI creates new opportunities for employees: dangerous, strenuous and repetitive work could be reduced and AI systems could support people in solving complex tasks. According to the report, however, a fine line exists between supporting people in their activities and restricting their freedom to make decisions themselves.
The study commission remains guarded on whether, as is widely discussed in the media, the use of AI systems endangers jobs. There is a dearth of reliable research results on this subject. The report concludes that past phases of automation did not result in large net decreases in employment, owing to the compensating effects of growth. New uses of AI however are now also affecting areas of cognitively demanding work that in the past proved relatively resistant to automation. AI could therefore contribute to resolving the shortage of skilled labour. It is conceivable that job quality will be enhanced and that the automated activities are in any case those presenting difficulties for human labour or giving rise to adverse working conditions. It is also possible, however, that a disparity could arise on the labour market if certain jobs are eliminated but skilled workers are not available in sufficient numbers for the new jobs being created in the field of AI.
The report concludes that the use of self-learning machines also has an impact upon the protection of individuals‘ rights to control their own data, the organization of co-determination, transparency and cultures of trust. Users of AI on both the employers‘ and employees‘ sides, and among their associations and trade unions, should therefore – in accordance with the principle of good work by design – be able to contribute effectively to defining the goals and configuration of AI systems, just as they contribute to their evaluation and operation and to further development of the social and technical conditions under which such systems are used. The commission calls for easy access to further training and advice on extending the AI expertise of those concerned. Ultimately, people will have to take decisions, for example regarding personnel, an area that is increasingly being partly or fully automated by AI-based systems.
In 2021, the European Commission presented a draft regulation governing artificial intelligence, incorporating the various national AI strategies of the EU Member States. The document makes provision for rules of conduct, minimum standards and prohibitions for four risk levels, with the intention of ensuring user safety and increasing confidence in the development and diffusion of AI. The Commission‘s aim here is to turn Europe „into the global hub for trustworthy Artificial Intelligence (AI)“.
Stefan Krempl
(freelance journalist)