What The Gufo published in Social Sciences this week.
ChatGPT users often prefer original prompts to AI reformulations
A user study on The Gufo's radar finds that GPT-4 reads common intents better than GPT-3.5, but reformulating prompts does not always please users more.
Consent needs scaffolding, not just information
A paper in the Journal of Medical Ethics argues that informed consent should be understood as something clinicians help build, not a single moment of disclosure.
New in the library
- User intent recognition and satisfaction with large language models: A user study with ChatGPT — Computers in Human Behavior: Artificial Humans
- Scaffolding informed consent — Journal of Medical Ethics
- LLM-powered socially assistive robot-delivered cognitive behavioral therapy exercises: an exploratory study with university students — Frontiers in Robotics and AI
- What is the Role of Small Models in the LLM Era: A Survey — Computational Linguistics
- TOP 2025: An update to the Transparency and Openness Promotion Guidelines — Research Integrity and Peer Review
- International heritage law as resistance to occupation — Leiden Journal of International Law
