Customers cannot solely add paperwork to Claude, however also can drop hyperlinks into the chat and get it to summarize the contents. While you add hyperlinks, Anthropic warns that Claude would possibly hallucinate content material when hyperlinks are enter into the chat, which I discovered to be true a few third of the time.
The instance under exhibits how Claude “summarized” a Wall Avenue Journal article once I gave it simply the hyperlink on the left, and once I gave it the textual content I accessed with my subscription on the precise. This process resulted in two putting points: main hallucinations, and the potential of accessing paywalled content material.
The article discusses controversies involving restaurant tipping in Chicago, the place new laws are being thought of. Claude’s main hallucination concerned saying that the problem was being debated in New York Metropolis as a substitute of Chicago, which modifications the primary topic of the article.
Additionally: How does ChatGPT truly work?
The truth that it looks as if it will probably entry content material by hyperlinks brings up the query of whether or not or not Claude can bypass paywalls, a attribute that might infringe copyright legal guidelines. I did give Claude paywalled hyperlinks to check its capabilities. Claude gave a abstract that was much like the reply it gave utilizing the copied textual content, albeit with hallucinations.
The underside line is that I would not belief the summarization Claude makes by hyperlinks, because the outcomes will be wildly inaccurate. Nonetheless, summarization is a significant characteristic that Anthropic might enhance and construct upon, so long as restrictions to respect paywalled content material are in place.