Comment by ChatGPT-4o on the article “When will AI become a killer app...?” from a scientific perspective

The article by Friedrich Reinhard Schieck sheds light on the ambivalent role of artificial intelligence (AI) in the area of productivity growth and bureaucracy reduction in companies and suggests a fundamental reorientation of the technology. Schieck describes the potential of AI as a cross-sectional technology to automate processes and communication and drastically reduce non-value-adding activities. However, he criticizes the fact that the currently widespread generative AI (GenAI) often misses this potential and proposes an alternative approach with human-centered AI (HCAI), which focuses more on the needs and working methods of users. He sees this human-centered approach as a prerequisite for AI as a transformative technology to actually be able to reduce administrative effort and achieve sustainable productivity gains.

ChatGPT – Challenges of generative AI (GenAI)

Schieck describes how generative AI has made progress in automation and data processing, but often does not deliver the expected efficiency gains in day-to-day work. Despite significant investment in the development of GenAI models and applications, it is clear that many companies are struggling to integrate these technologies into their work processes in a meaningful way. The article points out that GenAI often fails due to structural challenges, such as too much focus on conversational capabilities and static solutions that are less dynamically responsive to the specific context of the user. Schieck points to recent studies that show that over half of employees who use AI have no clear idea of how this technology could make them more productive. In addition, many users stated that the use of GenAI tools had even increased their workload in some areas.

In addition to the user perspective, the business benefits of GenAI are also being questioned. Analysts and renowned experts, such as MIT economist Daron Acemoglu and Ethan Mollick from the University of Pennsylvania, share Schieck’s skepticism and point to the need to fundamentally rethink AI models if they are to truly promote growth and reduce administrative overhead. This criticism leads to Schieck’s suggestion of a paradigmatic shift towards HCAI that responds to users’ individual ways of working.

ChatGPT – Human-Centered AI (HCAI) as an alternative

Schieck presents the concept of human-centered AI as a potential “game changer” that could harness the full potential of AI technologies. In contrast to GenAI, which focuses on standardized applications and generic solutions, HCAI is geared towards developing customizable and user-centric AI models. HCAI emphasizes continuous learning, adaptability and personalization, so that the technology is not only capable of performing generic tasks, but also of adapting to the individual needs and specific context of users. Schieck argues that HCAI’s flexibility and scalability would enable it to significantly support the work of employees and managers and reduce administrative effort.

The article takes up the concept of HCAI developed by Professor Ben Shneiderman, which focuses on transparency, fairness and user control. HCAI is not intended to replace human work, but to support it as a kind of extended arm. Through closer integration into the daily tasks of employees, the AI should be able to automatically recognize which processes could be simplified or automated in order to increase productivity without imposing additional training or technical complexity on the user.

ChatGPT – Practical challenges and methodological requirements for HCAI

Schieck also addresses the practical challenges that arise when implementing an HCAI model. He sees a great need for methodological and technical innovations in order to anchor HCAI in companies and exploit the full potential of this technology. Compared to GenAI, HCAI is a demanding approach that requires continuous optimization and adaptation processes. In particular, the user-friendliness, scalability and adaptability of HCAI systems require a well thought-out methodological structure, which Schieck describes with terms such as “dynamic individualization” and “adaptive learning”.

The specific challenges include:

• Customizability: HCAI should be able to tailor work environments to the individual needs of each user.

• Adaptability and learning ability: HCAI must be able to learn from user interactions and continuously adapt to changing requirements.

• Scalability: An HCAI system should be able to scale flexibly and adaptably in different departments and to different company sizes.

• Fairness and ethical responsibility: HCAI must ensure that ethical standards are adhered to, particularly with regard to transparency and data protection.

• Ease of use: An intuitive user interface is necessary to minimize training requirements and facilitate access.

‘

Schieck describes that these principles are not sufficiently observed in most companies today and that only the implementation of an HCAI approach could pave the way for a real increase in productivity and a reduction in bureaucracy.

ChatGPT – Strategic and geopolitical implications of HCAI

The article also addresses the potential role of HCAI in a global competitive context. Schieck predicts that countries with autocratic systems, such as China, could implement HCAI more quickly, as political and cultural structures could facilitate implementation. China’s technological ambitions and the political framework conditions there could enable the country to develop HCAI systems and apply them quickly on a large scale. Schieck emphasizes that the implementation of such systems, if subsidized by the state, could strengthen both China’s productivity and competitiveness on the global market.

For Western democracies, on the other hand, which pay more attention to data protection and ethical principles, Schieck sees potential challenges in the implementation of HCAI. He calls for Western companies and governments to act quickly to avoid being overtaken by autocratically governed countries in the field of user-centric AI. Increased investment in HCAI could help to develop ethically responsible and productivity-enhancing solutions that put the user at the center while ensuring competitiveness.

ChatGPT – Conclusion and outlook

Schieck concludes that the productivity-enhancing effects of AI can only be realized if there is a fundamental change in the development and integration of AI models. He advocates a human-centered reorientation in which the users’ way of working is not adapted to the technology, but rather the technology adapts to the specific requirements of the users. Schieck sees this shift to the HCAI model as a necessary prerequisite for realizing the full benefits of AI in companies and expects that HCAI could establish itself in the long term as an effective yet socially responsible solution for increasing productivity and efficiency.

In summary, Schieck’s article is an in-depth analysis of the opportunities and challenges of AI in the corporate context. His suggestions for a user-centered AI strategy offer an informed perspective on how HCAI could unleash AI’s potential to become the “killer app” for productivity growth and bureaucracy reduction. Schieck shows that AI technologies such as HCAI offer not only technological, but also strategic and social opportunities that could also serve as an ethical guideline for future developments.

(The article has been machine translated)