Aisvarya Chandrasekar
@aisvarya17.bsky.social
I work at @towcenter.bsky.social as a computational research fellow, studying the impact of AI tools in journalism. I am a grantee in Brown Institute’s 2024-25 Magic Grant cohort, building PollFinder.ai.
@technologyreview.com blocks all of OpenAI's and Perplexity's disclosed web crawlers in their robots.txt. Despite this, both OpenAI’s Atlas and Perplexity’s Comet retrieved the full text of a subscriber-exclusive article when asked to. How?
Read @klaudia.bsky.social and my recent CJR article:
Read @klaudia.bsky.social and my recent CJR article:
October 30, 2025 at 3:07 PM
@technologyreview.com blocks all of OpenAI's and Perplexity's disclosed web crawlers in their robots.txt. Despite this, both OpenAI’s Atlas and Perplexity’s Comet retrieved the full text of a subscriber-exclusive article when asked to. How?
Read @klaudia.bsky.social and my recent CJR article:
Read @klaudia.bsky.social and my recent CJR article:
While getting the ‘right’ answer from a chatbot is certainly very important, it is not always a marker of preferred behavior, especially when it suggests a potential violation of restrictions placed by publishers.
March 13, 2025 at 6:40 PM
While getting the ‘right’ answer from a chatbot is certainly very important, it is not always a marker of preferred behavior, especially when it suggests a potential violation of restrictions placed by publishers.
yes absolutely - the two pro models that we tested were Perplexity Pro and Grok-3 (beta). In the scope of our study, we didn't observe Perplexity Pro hallucinating URLs, but Grok-3 certainly did.
March 6, 2025 at 6:35 PM
yes absolutely - the two pro models that we tested were Perplexity Pro and Grok-3 (beta). In the scope of our study, we didn't observe Perplexity Pro hallucinating URLs, but Grok-3 certainly did.