Explore our podcasts to learn more about our research findings
-
Consider the following two situations: you notice a stranger struggling to carry a large box and offer help; you then share the physical effort to transport it to a goal location. Later at home, you and a partner assemble a jigsaw puzzle, coordinating your actions toward a shared goal. What distinguishes cooperation from coordination in these everyday interactions? This podcast summarizes our paper, in which we use the Rabbit-Duck illusion to illustrate real-life problems from the ambiguity between cooperation and coordination in AI systems. (We used Google’s notebookLM to generate the podcast out of our paper.)
- In emotion-rich contexts, how do you comprehend the meaning behind your perception? How do you play with connotation to create hidden messages that others can understand? This cool podcast summarizes our research, “Connotation and 3D Modeling from Limited, Raw Textual Descriptions“, published in the International Journal on Advances in Life Sciences. The podcast was created using Google NotebookLM, and although I would certainly flip the voices, it is so much fun!
2. Can a simple agent design produce complex behaviors, potentially leading to anthropomorphic interpretations? We adapted Braitenberg’s vehicles in a robot simulator to explore the distinction between embedded behaviors (the actual agent architecture) and interpreted behaviors (how we may perceive the agent’s actions). Observers often project more complexity onto AI than exists, influenced by context and anthropomorphic language, which can lead to a misunderstanding of the AI’s true capabilities and goals. Understanding the distinction between behavior and design is crucial for AI literacy. Podcast generated by NotebookLM from our paper” Complex Behavior Vs. Design -Interpreting AI: Reminders from Synthetic Psychology”.