Beskrivning:
"While capabilities of the modern AI/ML-based systems are impressive, there are a lot of questions related to responsible development and deployment of such systems in organizations, as well as ethics of using personal data. Such questions are especially poignant in cases where, so called, black box AI is being used. During this seminar we will discuss contemporary challenges of AI and personal data in organizations, and potential approaches to developing and deploying responsible and sustainable AI systems. We will present recent research and cases from public and private organizations that have navigated challenges of AI and data in their deployments."
Seminariet kommer att hållas på engelska.
Program:
09:00 – 09:05 Introduction
Aleksandre Asatiani, University of Gothenburg
09:05 – 10:00 AI - a prerequisite for a sustainable future and why we need to ensure trust when using AI
Patrick Eckemo, Chief Digitalisation & Innovation Officer, Bolagsverket; Anders Lagerqvist, Senior Architect, Lantmäteriet.
10:00-10:30 Coffee (included)
10:30 – 11:15 Is explainable AI the Holy Grail of AI?
Aleksandre Asatiani, University of Gothenburg
Mainstream popularity of AI-enabled organizational systems, the topic of explainable AI (XAI) has re-emerged. XAI strives to develop AI systems that offer transparent decision-making process and provide outputs that are interpretable and explainable to the stakeholders of the system. This presentation offers an alternative to XAI and discusses the case of successful development and deployment of the ‘black box’ AI systems at the Danish Business Authority.
11:15 – 11:45 Responsible data use
Olgerta Tona, University of Gothenburg
When companies manage employee data responsibly, they’re better able to grow trust while gaining insights. Putting dignity at the center of acceptable data use not only improves management of employee data but also allows transformation leaders to reinforce the organization’s value for and commitment to its employees.
