YouTube is testing a feature that allows users to flag suspicious and low-quality AI-generated videos, which could affect their visibility and monetization. This new rating system complements the standard moderation tools and helps the platform combat poor-quality content.
An international team of experts has developed a new exam, Humanity's Last Exam, designed to objectively assess the capabilities of artificial intelligence. The test includes tasks that current AI models are still unable to solve. Initial results show that even the most advanced AI systems are far from reaching human-level performance, and human expertise remains irreplaceable.
A new study has shown that AI chatbots designed to support mental health often violate professional and ethical standards, even when provided with specific guidelines. Experts emphasize the need for rigorous evaluation and regulation of such systems before they are used in critical situations.
Crius
