A new paper dissects flaws in AI safety discourse around open weights. It argues current debates ignore practical trade-offs and ideological biases, calling for more nuanced risk assessments. Critical for AI developers balancing openness, safety, and real-world deployment challenges.
Opening Kapyn…