AI can beat chess masters but PDFs remain its final boss

A tech news graphic featuring a blue robot head holding orange PDF files with question marks, symbolizing AI's difficulty processing PDFs. The image illustrates a report from The Verge about how modern AI systems struggle with PDF comprehension despite advanced capabilities in other areas. Extracted text: A new report from The Verge says modern AI still has trouble reading PDFs. Even though AI can write code and answer complex questions, it often struggles to pull clear and accurate information from PDF files. The problem became clear when developers tried to search through millions of Justice Department documents released as PDFs. Many files were poorly scanned and hard to search, even after text recognition was added. Since PDFs are built to look good on screen rather than store clean data, AI finds it difficult to understand tables, emails, and layouts properly. This highlights a gap between AI's technical ability and real-world reliability. Follow us (@therundownAI) to keep up with the latest news in tech and AI. Source: The Verge "How many AIs does it take to read a PDF?"z AI NEWS AI still struggles to reliably read PDFs despite major progress