Companies with a future: developing AI skills
Leadership and action areas for building artificial intelligence skills
AI is fundamentally changing the economy and presenting new opportunities for automation, increasing efficiency and personalising products. To translate these opportunities into competitive advantages, but also limit the risks associated with AI, there is a need for managers who have the required skills to successfully support this process and competitively position themselves for the future. This is not only necessary from an economic point of view, so that Germany is favourably positioned among the international competition; corresponding measures also need to be implemented from an individual business perspective.
Definition of an AI manager
An AI manager is a person who holds a leadership position within a company or organisation and who is responsible for the development, implementation and monitoring of artificial intelligence (AI) within a company, i.e. in its products or processes.
Artificial intelligence – where do we currently stand?
As AI grows in importance as a key technology, many countries have developed AI strategies or position papers to actively shape the associated political and social challenges and developments. Despite differences in terms of their specific design, the strategies are similar: not only are significant resources being invested in AI research, but the application and expansion of the technology is also being supported in a targeted manner, for example in key sectors such as smart cities, the military and the medical sector.
With a view to tapping into the expected economic growth of 430 billion euros 1 (PWC (2019a): Aus dem Hype Realität machen: Fit für Künstliche Intelligenz im Jahr 2020 [Living up to the hype: Fit for artificial intelligence in 2020]), the German government has already begun implementing measures: it has provided over 2.5 billion euros in funding for AI initiatives since 2019, of which more than 490 million euros has already been spent on projects, most of which will run over multiple years2.
Germany published3 its AI strategy in November 2018, trailing behind countries such as the USA (2016), China (2017) and Canada (2017). Nevertheless, AI research is well and widely established in Germany, even by international standards. The assumption that Germany ranks among the top 5 in terms of international competition in this field is supported by the number of German AI publications, as well as the frequency with which German and European AI scientists are cited4(Jahn, Thomas; Schinroszik, Nadine (2023): Was Start-ups und KI-Forschung in Deutschland bremst [What is hindering startups and AI research in Germany] In: Handelsblatt, 12 April 2023). But despite this strength on the AI research side, analysts note that Germany is falling behind in terms of using AI in the economy, is weak in terms of innovation and is failing to translate the comparatively high R&D expenditure into value creation. According to a Bitkom survey, just nine percent of all companies are currently using AI5. Other analyses, which take into account the reasons Germany is lagging behind, have reached similar conclusions. For example, findings indicate that German managers view AI as less important than the international average6. This appears to be impacting its application: even where AI is introduced, it is primarily used as a means of improving productivity through automation, and less in the area of innovative products and business models 7 (see PWC (2019), p.5 et seqq.). Disruptive, radical innovations that are based on AI and which could fundamentally change existing markets – or even create new markets – and offer long-term competitive advantages are hardly being promoted at all.
It can be assumed that there are various reasons for this. Possible causes include the lack of technical AI infrastructures, high levels of regulation and high implementation costs. However, the lack of AI talent and appropriately qualified AI managers who have the skills needed for the AI revolution is probably also a contributing factor.
Side note: AI research & application must go hand in hand
According to the definition by John McCarthy, who coined the term, AI is the science and engineering of making intelligent machines. AI research must be firmly anchored within the IT sector, even though it also intersects with many different scientific domains and research disciplines. Areas of application for AI include
- industrial use (e.g. robotics, error checking and quality control in production, logistics planning and supply management),
- the financial sector (e.g. exposing fraud, risk analysis for loans and mortgages, predicting share prices or the valuation of companies),
- environmental monitoring (e.g. monitoring water quality and predicting and identifying pollution), mobility (e.g. self-driving cars, driver assistance systems),
- medicine and pharmaceuticals (e.g. diagnostic support, active substance development),
- applications in cross-disciplinary functions such as marketing (e.g. analyses) or design.
At the moment, the media focus is on large AI models such as the large language model ChatGPT. These large models give people an idea of what AI could be used for in the future. However, they are generally not suitable for complex tasks. There is still a significant need for further research. Existing AI systems are predominantly either expert systems, which can derive recommended actions and support solutions to complex problems but cannot learn new rules or identify facts, or they are statistical learning programs, which automatically identify patterns for concrete problems on the basis of statistical probabilities using large quantities of data.
Much research and development is still needed to expand the possibilities of using AI for new applications, lower the requirements and facilitate large-scale implementation. The research needed ranges from algorithmic foundations to AI systems that will simplify the application of AI. After all, AI has been such a complex topic that so far only a small number of companies with specialised teams have been able to translate it into practical applications.
These AI skills need managers
AI offers huge potential for the economy, and companies need to be prepared for it. Although many of the technologies that fall under the broad term of AI are still in the early stages of development, companies are faced with the challenge of systematically, purposefully and profitably implementing AI in products or processes. This means that it is necessary not just to develop the required skills for the technical implementation of AI and support corresponding products – we also need to identify use cases that offer added value, take account of ethical considerations, minimise risks and support change processes. Depending on the positioning of the hiring company, the related tasks may be rooted in different dimensions of technology, management skills and leadership skills.
An AI manager role may include, among other responsibilities, identifying business opportunities for AI technologies, selecting the right technologies for specific business requirements, implementing AI solutions, monitoring AI systems, identifying risks and implementing risk mitigation measures as well as training employees in the area of AI. At the same time, to ensure that the AI systems are used effectively to achieve the business goals, but will also support the ongoing transformation processes. an AI manager must also have skills in the areas of management, communications, strategic thinking and problem-solving. In view of the current skills shortage, the AI manager’s responsibilities may also include managing product development service providers. In addition, an AI manager should have a deep understanding of ethics and data protection to ensure that AI systems are operated in accordance with the relevant ethical and legal standards.
Managers must boast a wide range of characteristics and qualifications in order to fulfil these responsibilities:
Technical expertise: An AI manager must have a profound understanding of AI technologies. They should be able to guide their team in selecting the right algorithms and techniques for solving specific problems. In light of the highly dynamic nature of the AI sector, managers also need to be willing to learn new things and must have access to good networks that enable them to continuously develop their skills.