AI as Teammate or Tool? A Review of Human-AI Interaction in Decision Support
A new paper analyzes 4 key dimensions of human-AI interaction, revealing a critical design flaw.
Deep Dive
Researchers from a Bangladeshi university published a review titled 'AI as Teammate or Tool?' analyzing Human-AI Interaction (HAI) literature. They found current AI systems remain largely passive due to overreliance on explainability-centric designs. The study identifies static interfaces and miscalibrated trust as key limitations. To become true teammates, AI needs adaptive, context-aware interactions that support shared mental models and dynamic negotiation of authority with humans.
Why It Matters
This challenges the core design of current AI assistants, pushing for systems that actively collaborate rather than just respond.