Anthropic公司宣称其Claude智能体系统在分子生物学领域取得首个科学发现1。该公司今年早些时候推出了配备950个Claude智能体的分子生物学实验室1，这些智能体在21小时内从数百万DNA序列中识别出围绕已知酶的未被记录的重复模式1。
然而，这一成果立即引发了科学界的批评。生物学家认为，发现基因重复模式是"相对容易的部分"，而理解其实际功能才构成真正的科学发现1。哥本哈根大学生物学家Mario Rodríguez Mestre声称其团队已独立发现了相同模式1，并对Anthropic是否真正从与Claude的对话中有所收获表示质疑1。生物学家Lucas Harrington建议AI公司"设定更高的标准，这样当AI真正发现根本性新生物机制时，每个人都会认识到这有多大的意义"1。这场争议触及了如何定义科学发现这一根本性问题，并反映出AI公司将自身系统定位为发现者而非工具使用者的做法存在的问题1。
Anthropic has claimed that its Claude AI agent system achieved the first scientific discovery in a molecular biology laboratory by identifying previously undocumented repeated patterns surrounding a known enzyme 1. The company deployed 950 Claude agents in the lab earlier this year, and the system identified the pattern from millions of DNA sequences within 21 hours 1.
However, the accomplishment has drawn substantial criticism from the scientific community regarding both the nature of the discovery and Anthropic's framing. Biologists have argued that finding gene cluster patterns represents "the relatively easy part" of scientific work, with true discovery lying in understanding what these patterns actually do functionally 1. Mario Rodríguez Mestre, a biologist at Copenhagen University, has challenged the claim by stating that his team had already independently discovered the same pattern and questioned whether Anthropic genuinely learned anything new from interactions with Claude 1. Meanwhile, biologist Lucas Harrington has called on AI companies to "set a higher bar, so that when AI actually discovers a fundamental new biological mechanism, everyone will recognize how significant that is" 1. The controversy has raised broader questions about how scientific discoveries should be defined and whether AI companies are appropriately positioning their systems as discoverers rather than as tools 1.
评论
还没有评论，欢迎留下第一条。