This Week in AI: Generative AI is spamming up academic journals
How do the fake journals discussed in the article manipulate the CiteScore system to achieve high rankings?
The fake journals discussed in the article manipulate the CiteScore system by extensively cross-citing each other, which boosts their citation count and consequently their CiteScore ranking. This tactic takes advantage of the fact that CiteScore considers citations in its calculations, allowing the journals to rank highly despite being comprised of AI-generated articles with little to no actual scholarly value.
How is DeepMind's new AI technology designed to generate soundtracks for videos described in the article?
DeepMind's new AI technology, Video-to-Audio (V2A), is designed to generate soundtracks, sound effects, and dialogue for videos. It uses video pixels and optional text prompts to create synchronized audio that matches the visual content of the video. The tool can generate an unlimited number of soundtracks for any given video, providing a vast array of audio options. Users can provide positive or negative prompts to refine the audio output, offering greater creative control. This technology has the potential to revolutionize how soundtracks and audio effects are created and integrated into videos.
What specific issues have arisen from generative AI's involvement in academic publishing as highlighted in the Retraction Watch blog post?
Generative AI's involvement in academic publishing has led to the creation of fake journals, filled with AI-generated articles. These journals game the systems used to evaluate researchers for promotions and hiring, as they extensively cross-cite each other to inflate their rankings in metrics like CiteScore. This could potentially influence academic awards, hiring, and promotion decisions, based on the perceived quality of research that is, in fact, AI-generated spam content.