Anthropic 发布了一套名为 Model Hardware Standard (MHS) 的标准化驱动集合，旨在让人工智能代理能够轻松连接并控制各类物理设备 1。该标准目前处于“研究预览”阶段，主要面向科学家群体，用于简化实验组件的定制软件集成过程 1。Anthropic 指出，MHS 能够将原本需要数周或数月才能完成的实验搭建时间缩短至数小时或数分钟 1。
在技术层面，MHS 提供了通用的接口与数据格式，使得实验中的不同设备无需编写定制的“翻译”程序即可实现跨网络通信 1。Anthropic 技术人员 Alek Kemeny 表示，这一想法源于他在位于弗吉尼亚州阿什本（Ashburn）的 HHMI Janelia 研究园区观察神经科学家 Arco Bast 进行脑记忆形成实验时的经历 1。在实验中，Bast 需要协调激光、显微镜和相机等多种组件，这启发了团队开发 MHS 以解决设备协同难题 1。
Anthropic has released the Model Hardware Standard (MHS), a collection of standardized drivers designed to enable AI agents to easily connect with and control various physical devices 1. Currently in a research preview phase, MHS is primarily aimed at scientists to simplify the integration of custom software for experimental components 1. By providing a universal interface and data format, MHS allows different devices to communicate across networks without the need for custom translation programs 1. According to Anthropic, the system can reduce experiment setup times that typically take weeks or months down to hours or minutes 1.
The idea for MHS originated from observing neuroscientist Arco Bast's brain memory formation experiments 1. Anthropic Technical Staffer Alek Kemeny stated that the concept was inspired by watching Bast coordinate experimental components such as lasers, microscopes, and cameras 1. These observations took place at the HHMI Janelia Research Campus, located in Ashburn, Virginia 1.
评论
还没有评论，欢迎留下第一条。