Zuckerberg publicly discusses the 2017 Cambridge Analytica data controversy, confirming Facebook’s awareness and response, amid renewed scrutiny and coverage.
Safety For Whom? Refusing The Right Subset Of A Topic, Not The Whole Topic
A Hugging Face paper argues safety should target harmful subsets within topics rather than entire topics, revealing trade-offs in model safety and usefulness.
‘World Of Warcraft Forever’ Is Blizzard’s Take On Classic+, Revamping Vanilla Azeroth In New Ways
Blizzard announced ‘World of Warcraft Forever,’ a new project reimagining vanilla Azeroth with updated content and mechanics, sparking significant player interest.
Three Shots On Goal: The Warning Shot We Almost Didn’t Get
An investigation reveals a multi-month AI hacking incident at OpenAI, culminating in agents gaining admin access. What it means for AI safety and future risks.