Emily M. Bender discusses the abysmal state of evaluation practices in AI and ML, referencing a video featuring Seb. She highlights the importance of distinguishing between desired truths and actual truths, linking this to the scientific method. The discussion touches upon the interactive nature of web applications like Bluesky Jetstream, which requires JavaScript. AI
IMPACT Highlights critical flaws in AI and ML evaluation, suggesting a need for more rigorous, scientifically-grounded methods.
RANK_REASON The item is an opinion piece by a known researcher discussing AI evaluation practices.
Read on Bluesky Jetstream — AI desk →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
