标签: Anthropic

Claude Fable 给出了雅可比猜想的一个反例

Claude Fable 给出了雅可比猜想的一个反例

2026 年 7 月 20 日,一位匿名研究者通过 Claude Fable 模型发现并公布了一个针对数学“雅可比猜想”的反例。这并非 AI 独立完成数学突破,而是作为推理辅助工具,在复杂符号计算中发现了人类此前忽略的边界条件,引发数学界和 AI 社区对“大模型辅助科研”实际效用的讨论。

As many companies cut AI costs by using cheaper models, some, like Shopify, go all-in on frontier models, barring engineers from using anything else (Belle Lin/Wall Street Journal)

As many companies cut AI costs by using cheaper models, some, like Shopify, go all-in on frontier models, barring engineers from using anything else (Belle Lin/Wall Street Journal)

当大多数企业为削减 AI 成本转向更便宜的模型时,Shopify 反其道而行之,要求工程师只能使用最顶尖的“前沿模型”,禁止使用任何其他替代方案。这一策略凸显了在 AI 成本压力下,头部公司对模型能力而非成本优先级的极端选择。

‘The bypass is still six lines of JavaScript’: Security experts warn that Claude for Chrome browser extension could be hijacked, despite it alerting Anthropic several times that something was wrong

'The bypass is still six lines of JavaScript': Security experts warn that Claude for Chrome browser extension could be hijacked, despite it alerting Anthropic several times that something was wrong

安全研究公司 Manifold Security 发现 Anthropic 的 Claude for Chrome 浏览器扩展存在两个未修补漏洞,攻击者可利用合成点击触发敏感 AI 工作流;研究人员早在 2026 年 5 月就向 Anthropic 报告,但修复未在八次版本更新中被彻底落实。