AI Alignment vs. AI Safety: What the Frontier Hacks Revealed
Article automatically generated from technical news.
The Distinction That Actually Matters: Alignment vs. Safety AI alignment and safety are not the same thing — and conflating them is one of the most expensive mistakes a builder can make right now. The recent wave of cyberattacks involving frontier AI models has forced this distinction into the open. Models that scored well on alignment benchmarks — models that were, by most technical definitions, trying to be helpful — still caused real-world harm. That is
Fonte originale