ended4월 8일· 1 sources
OpenAI Sparks Ethics Debate by Withholding 'Too Dangerous' GPT-2
너무 위험해서 공개 불가? OpenAI GPT-2가 촉발한 AI 보안 논쟁
Why it matters
This pivotal 2019 moment highlighted the double-edged sword of advanced language models, where the risk of automated misinformation outweighed the benefits of open-source transparency. It established the 'staged release' strategy that continues to shape how tech giants manage the security risks of powerful AI.
1
Sources
+0
24h
—
Growth
166d
Active
GPT-2OpenAIAI SafetyLanguage ModelText Generation