Cointime update: OpenAI’s chief scientist said on Nov. 23 that the company will increase transparency around its pre-training progress to help build shared AI safety benchmarks. He proposed setting unified limits based on metrics such as perplexity, with any breach conditional on first having the ability to understand AI behavior. He also predicted a wave of technical breakthroughs in AI pre-training before 2027.
OpenAI Chief Scientist Pledges More Transparency on Pre-Training Progress for AI Safety Benchmarks
Share on WeChat
Open WeChat → Scan → then tap "…" to send to a chat or Moments.
Tap "…" in the top-right corner to send to a chat or share to Moments.
Related News
7h ago
MiniMax Enters AI4S Research, Publishes Cryptanalysis Findings
7h ago
Tencent Opens Limited Beta for AI Voice Input Tool Chatterfly
9h ago
SpaceX, Anthropic IPOs Hand Underwriters $500M+ Each Amid Process Complaints
14h ago
US Intelligence Chief Clayton: Nvidia Investigation Will Proceed Naturally
15h ago
Musk warns AI industry may face severe funding shortfall next year
15h ago
MiniMax Internal Model Breaks NIST Cipher Candidates in Automated Cryptanalysis