AIForest

À Propos

AI Forest is an innovative interactive game and research environment designed by the LIT Robopsychology Lab and the Visual Data Science Lab at JKU Linz, focusing on human-AI decision-making processes and the explainability of AI systems. Its core purpose is to serve as a scientific tool for studying how humans interact with and trust AI, especially in scenarios requiring decision-making based on machine-generated explanations. Players participate by navigating an indoor woodland environment, searching for physical mushroom objects, and using a tablet equipped with an AI-powered app to scan and classify the fungi as either edible or inedible. This practical interaction generates vital data on how users understand and rely on AI-led decisions, offering insights into effective explanation strategies across various demographics. A key feature of AI Forest is its engaging gameplay, which blends physical exploration with AI technology, making the entire experience immersive. The game collects metadata about player interactions, enriching research on human-AI collaboration. The app's ability to explain its classification decisions is central to the research, hence prioritizing explainability and transparency. While the current application is educational and research-focused, AI Forest opens possibilities for adaptation into other fields involving human-AI interaction. Potential future implementations could include healthcare diagnostics, financial systems, or environmental monitoring, where trust and transparency in AI systems are crucial. AI Forest stands out due to its unique combination of physical gameplay with a focus on AI explainability, broadening the scope for understanding AI trust dynamics in real-world scenarios. Although specific technical details of its algorithms and data infrastructure are not specified, the development team indicates the AI has been trained extensively on mushroom imagery. Currently, the tool operates as a self-contained system without specified integration capabilities. However, future versions could benefit from integrations with broader datasets or platforms, potentially expanding their educational and research applications. As of the latest available information, AI Forest's recent developments or achievements are not detailed, suggesting a need for further updates or direct inquiry for the latest project progress. Regardless, AI Forest remains a compelling initiative in the realm of interactive AI research, aiming to deepen our understanding of human-AI interaction and build more trustworthy, explainable AI systems.