Archives
The cuttlefish.org archive is the basis of Agent Cuttlefish's historic knowledge. It contains websites and documents dating back to 1993, when Sean Clark created his first website. When fully uploaded, it will contain over 50,000 items. These include photographs, web pages, graphics, videos and sounds.
The challenge working with this amount of data is not simply its size, but extracting useful knowledge from it. It can be fed in to an AI system, but my experiments have shown that human-added metadata (such as dates, locations and background) greatly enhances the materials. As such, maintaining and developing the archive is an ongoing and interactive process.
This is still a work in progress — some features may not work.