Anthropic称新防护使竞品蒸馏Claude效率降低90-99%
Anthropic says new safeguards have cut the distillation that rival labs can pull…
Anthropic says new safeguards have cut the distillation that rival labs can pull out of Claude by 10 to 100 times ("around 90 to 99 percent").
Anthropic 表示，新的安全措施已将竞争对手从 Claude 中提取的蒸馏数据量减少了 10 到 100 倍（"约 90% 至 99%"）。
Anthropic names Chinese labs Zhipu and Moonshot, and says one network once ran more than 20,000 fake accounts at a time.
Anthropic 点名了中国实验室智谱和月之暗面，并表示其中一个网络曾同时运行超过 20,000 个虚假账户。
It remains to be seen whether GLM and Kimi will continue to release outstanding models when they can barely distill anymore.
目前尚不清楚 GLM 和 Kimi 在几乎无法再进行蒸馏的情况下，是否会继续发布出色的模型。
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力
