An AI tool using language models like BERT has flagged over 250,000 cancer research papers as suspected paper-mill products, revealing a steep rise in low-quality or fraudulent publications over the past two decades, though further verification by human experts is needed due to potential false positives.
Researchers at the University of Kansas have developed an AI text detector that can accurately distinguish between human-written and computer-generated content in scientific essays, specifically in the field of chemistry. The detector, trained on journals published by the American Chemical Society, achieved almost 100% accuracy in identifying human-authored passages and reports generated by AI models like ChatGPT. Existing AI detectors for general content performed poorly in detecting AI-generated content in scientific papers. The tool aims to help academic publishers assess the infiltration of AI-generated text, mitigate potential problems, and ensure the integrity of scientific literature.
Customer support agents using a Generative Pre-trained Transformer (GPT) AI tool saw a nearly 14 percent increase in productivity, with 35 percent improvements for the lowest skilled and least experienced workers, and zero or small negative effects on the most experienced/most able workers. The AI tool offered potential responses to customer queries, and agents could choose to take those suggestions or ignore them and enter their own responses. The tool helped newer and lower skilled workers adopt the skills and behavior of more experienced workers more quickly, and customers were more likely to express positive sentiments when interacting with agents using AI assistance.
Google has unveiled Project Tailwind, a prototype AI notebook that can be trained on users' documents and used as a personalized tutor or writing companion. The tool can be used by anyone who deals with a lot of text in their life, including writers, analysts, and lawyers. The AI notebook is powered by the user's notes and sources, and can create a personalized and private AI model that has expertise in the information provided. However, there are potential challenges such as compute demands and reliability. Google is inviting users to sign up to test Project Tailwind as part of its new AI Labs program.