YouTube is rolling out new settings that let content creators decide if AI companies can use their videos for model training. The new controls appear in YouTube Studio under "Third-party training," ...

The benchmark uses 1,719 selected examples where AI models must generate detailed responses based on provided documents. The benchmark's unique feature is its evaluation method: three leading AI ...

Meta and Stanford researchers have developed Apollo, a new family of AI models that tackles one of AI's persistent challenges: getting machines to truly understand videos. While AI has made huge ...

Google unveiled two new AI models today: Veo 2 for video generation and Imagen 3 for images. According to human evaluators, both models set new benchmarks in their respective fields. The new Veo 2 ...

According to Reuters, the United States is preparing new export regulations for AI chips that would position companies like Google and Microsoft as global gatekeepers for access to this technology ...

Midjourney released an early version of its new AI model customization system that lets users create multiple profiles and customize their AI models more quickly using collections of reference images.

During his presentation at the NeurIPS conference, AI pioneer and OpenAI co-founder Ilya Sutskever warned that AI development is approaching the limits of available training data. While computing ...

OpenAI brings an updated o1 model to its API, adding new features such as function calls, JSON-formatted structured output, and image analysis capabilities. The new model, dubbed "o1-2024-12-17," ...

Nvidia announced a new compact and affordable "supercomputer" for generative AI development, the Jetson Orin Nano Super. According to Nvidia, the developer kit delivers up to 70 percent more ...

Google Labs demonstrated new capabilities of its multimodal Gemini 2.0 Flash model, showing how users can now interact with images through simple taps and text prompts. The combination of visual and ...