Tag: Artificial Intelligence & Machine Learning

Study reveals human moderators favor powerful parties in asymmetric conflicts; AI assistance reduces most biases but sometimes amplifies them.

Fine-tuned language model provides automated question answering in large calculus courses, achieving 75% accuracy. Hybrid human-AI workflows maintain instructor oversight while scaling support.

Participatory research reveals gaps in deepfake sexual abuse moderation systems and proposes multi-stakeholder infrastructure improvements based on activist and survivor input.

Decentralized social media operators define boundaries for AI integration, rejecting autonomous systems in favor of tools that support human governance and community autonomy.

Raggy enables rapid experimentation in retrieval-augmented generation pipelines through composable primitives and interactive debugging, reducing development friction.

Panel discussion on generative AI’s multidimensional impact on HCI globally: education, industry dynamics, and funding across diverse regions and cultures.

DuoDrama uses AI agents to generate screenplay feedback that coordinates character perspectives with story-level views, improving screenwriter reflection during refinement.

Study on sonification design for hip-hop dance improvisation, examining how dancers experience real-time sound feedback during collective and individual movement performance.

Study of suppression-based visual filters (Dim and Blur) to enhance attention in dynamic virtual environments with competing objects. Results show Dim-Strong filtering achieves optimal performance.

Panel discussion on whether CHI should adopt formal reporting standards to improve research quality, transparency, and peer review processes, informed by experiences in adjacent fields.