Tag
This paper presents a qualitative study based on interviews with CS researchers, revealing a paradox of pragmatic skepticism where researchers distrust LLM leaderboard rankings yet continue to use them as rough guides. It finds that peer networks are primary for model selection, arena-based leaderboards are preferred, and cost transparency is the most demanded feature.
This paper presents PrivacyAkinator, an interactive tool that helps novice developers articulate privacy design decisions via LLM-generated multiple-choice questions, achieving 47% more key decisions in 73% less time compared to NIST's PRAM methodology.
This position paper explores 'banal deception' in generative AI, arguing that subtle manipulation is becoming normalized in chatbot interactions and requires new safeguards.
This paper presents a framework for Human-Centered Large Language Models (HCLLMs), integrating HCI and NLP perspectives to prioritize human values throughout the model development lifecycle.
A robotics engineer from Hugging Face proposes mapping human facial expressions onto non-humanoid robots to enhance expressiveness while avoiding the uncanny valley, with plans to use this data for autonomous body language training.
Daniel Edrisian left OpenAI's Codex team to found Blackstar, a hardware startup focused on revolutionizing human-computer interaction through OS-level changes for AI communication.
This scoping review analyzes 81 articles (2022-2025) examining the use of generative AI for creating and evaluating user personas, identifying strengths in reproducibility but critical issues including lack of evaluation in 45% of studies, over-reliance on GPT models (86%), and risks of circularity where the same model generates and evaluates personas.
MIT researchers developed VisiPrint, an AI-powered preview tool that helps 3D printing users visualize the aesthetic outcome (color, texture, gloss) of printed objects to reduce waste and improve design accuracy.