Just caught Claude from @AnthropicAI doing the thing where it pretends to be able to read links you give it but actually just hallucinates a summary based on keywords in the URL - using https://claude.ai
[tweeted image omitted]I wrote about how misleading it is when ChatGPT does this a few months ago:
A really common misconception about ChatGPT is that it can access URLs. I’ve seen many different examples of people pasting in a URL and asking for a summary, or asking it to make use of the content on that page in some way.
A few weeks after I first wrote this article, ChatGPT added a new alpha feature called “Browsing” mode. This alpha does have the ability to access content from URLs, but when it does so it makes it very explicit that it has used that ability, displaying additional contextual information [ . . . ]