谁在害怕中国模型？Ben Thompson 提议美国立法禁止服务条款限制蒸馏以应对 Qwen 等开源模型竞争

谁害怕中国模型？：https://stratechery.com/2026/whos-afraid-of-chinese-models/ （来源：https://daringfireball.net/linked/2026/07/20/thompson-chinese-models-distillation） 本·汤普森提出了一个有趣的建议，这一建议既指出了实验室在禁止对模型进行蒸馏时的虚伪（尽管他们使用未经许可的数据进行训练），也可能有助于美国开放模型更有效地与中国同行竞争：

本还推测，阿里巴巴决定发布 Qwen 3.8 Max 为开放权重——与他们在五月决定不发布 Qwen 3.7 Max（https://qwen.ai/blog?id=qwen3.7）相反——可能受到近期一次习近平演讲的影响：http://english.scio.gov.cn/topnews/2026-07/18/content_118605932.html，他表示：

我们应该抓住这个难得的历史机遇，鼓励开源、开放、协作与共享。

关于 Qwen 3.8 Max 的话题：https://twitter.com/Alibaba_Qwen/status/2078759124914098291 ——一个新的 2.4T 参数模型（几乎和 2.8T 的 Kimi K3 一样大）——这是它画的一只鹈鹕：https://tools.simonwillison.net/markdown-svg-renderer#url=https%3A%2F%2Fgist.github.com%2Fsimonw%2F735f2cf19b795517cb2ff6cae1c71c64

我特别喜欢在（详尽的）推理轨迹中看到这些注释：“可以加头盔吗？不。” 和 “也许加一个小铃铛？不。” 以及 “篮子里可能需要加一条小鱼？没必要。”

这是 Simon Willison 的一篇链接帖子，发表于 2026 年 7 月 20 日：/2026/Jul/20/。

每月赞助我 10 美元，即可获得经过策划的当月最重要的大语言模型发展电子邮件摘要。

Who’s Afraid of Chinese Models?：https://stratechery.com/2026/whos-afraid-of-chinese-models/ (via：https://daringfireball.net/linked/2026/07/20/thompson-chinese-models-distillation) Interesting proposal from Ben Thompson that both addresses the hypocrisy of labs outlawing distillation against their models despite training on unlicensed data, and could help US open models compete more effectively with their Chinese counterparts:

Ben also theorizes that Alibaba's decision to release Qwen 3.8 Max as open weights - a reversal from their decision not to release Qwen 3.7 Max：https://qwen.ai/blog?id=qwen3.7 in May - may have been influenced by a recent speech：http://english.scio.gov.cn/topnews/2026-07/18/content_118605932.html by Xi Jinping, who said:

We should seize this rare, historic opportunity to encourage open source, openness, collaboration and sharing.

And on the subject of Qwen 3.8 Max：https://twitter.com/Alibaba_Qwen/status/2078759124914098291 - a new 2.4T parameter model (nearly as large as the 2.8T Kimi K3) - here's a pelican it drew：https://tools.simonwillison.net/markdown-svg-renderer#url=https%3A%2F%2Fgist.github.com%2Fsimonw%2F735f2cf19b795517cb2ff6cae1c71c64:

I particularly enjoyed seeing these notes in the (extensive) reasoning trace: "Could add helmet? No." and "Maybe add small bell? no." and "Need maybe add small fish in basket? Not necessary."

This is a link post by Simon Willison, posted on 20th July 2026：/2026/Jul/20/.

Sponsor me for $10/month and get a curated email digest of the month's most important LLM developments.