The swift progress of artificial intelligence (AI) is profoundly shaping the world of work. Companies are challenged to create AI applications that are not only efficient but also trusted and embraced by those who interact with them. This article draws on contributions from Fraunhofer IAO, a pioneer in researching and developing human-centered AI applications.
What is human-centered AI?
Human-centered AI involves placing people at the core of AI application design and implementation. It focuses on crafting AI systems that consider and cater to the needs, abilities, and expectations of users.
Human-centered AI at FLOWIT
At FLOWIT, human-centered AI is central to our corporate philosophy. We prioritize developing AI applications that are not only efficient and powerful, but also intuitive and user-friendly. Our products and solutions are crafted in close collaboration with our customers to ensure they fully support their needs and workflows. Additionally, we minimize risks (e.g., anonymizing personal data) and maximize opportunities. FLOWIT collaborates interdisciplinary on advancing the Digital Coach. Experts from psychology, law, data science, and economics contribute to its development, reflecting a nuanced perspective on artificial intelligence.
Fraunhofer Institute Guidelines
At FLOWIT, we rely on the AI Test Catalog from the Fraunhofer Institute to shape our AI program for employee development, specifically focusing on non-desktop applications. This catalog provides a structured approach to identifying AI-specific risks, considering six dimensions of trustworthiness: fairness, autonomy and control, transparency, reliability, security, and privacy.
Our goal is to craft an AI program that not only emphasizes efficiency but also upholds the highest standards of trustworthiness and security. The AI Test Catalog helps us formulate specific test criteria for each AI application, ensuring they meet rigorous standards. It also guides the structured documentation of technical and organizational measures throughout our AI program's lifecycle.
These measures, reflecting the current state of the art, serve to mitigate potential risks associated with AI applications. By adhering to this approach, we ensure that our AI program not only delivers efficiency and effectiveness but also complies with ethical and legal standards. Our AI solutions are tailored to enhance employee development while meeting the needs of our customers and users.
Designing human-centered AI
Designing human-centered AI involves collaboration between various disciplines, such as AI experts, usability experts, and future users of the applications. Through iterative development, user needs can be continuously addressed, enhancing the acceptance of AI systems.
Human-centered AI will be pivotal in the future of work. Organizations developing and implementing AI applications should recognize the significance of the human element. By adhering to guidelines and engaging closely with users, they can create AI systems that are efficient, powerful, and appreciated by those who use them.
Source:
- Christian Knecht, Human-AI interfaces: The success factor in the introduction of artificial intelligence in production, Frauenhofer IAO
- Bastian Pokorni, Kickstarting the age of AI: how artificial intelligence can be successfully introduced in production, Frauenhofer IAO
- Martin Braun, Artificial intelligence and people: how manufacturing companies can strengthen their future viability, Frauenhofer IAO
