Discover explanable AI with the online game "Guess the City"
CS-15220
HigherEd
Universität Ulm
The XAI studio is the joint innovation laboratory of the Institute for Business Analytics at the University of Ulm and the start-up inlinity. In research and practical projects, new approaches to explainable AI are tested and the potential of this key technology for a responsible use of artificial intelligence is demonstrated.
The Challenge
Explanable AI (Explainable AI or Xai) has the potential to improve the handling of users with AI and to significantly advance the use of AI technologies. But how can the potential of explainable AI be made tangible and tangible? And how can it be examined how situation and target group-specific explanations have an impact on the use of AI?
The Solution
In teams from doctoral students and students from the University of Ulm and Experts of the start-up inlinity, the XAI demonstrator is developed-a web app that can be used to make methods of explainable AI can be experienced using simple application examples for the general public. The students get to know modern software development and how to deal with a state-of-the-art cloud environment.
The Results
The game "Guess the City" is one of the applications in the XAI demonstrator. Players guess where a street scene from Google Streetview was recorded. They compete against a AI that visually explains their decisions. Scientific evaluation shows that the players benefit from it. Some participating students have found great entry positions as data scientists.
The Highlight
We have set ourselves the goal of clearly conveying the topic that can be explained - both in teaching as well as in public and with possible users in practice. The development of vivid, easily accessible application examples is an integral part of it. For us, the Google Cloud Platform is the ideal place to implement these lighthouse projects.