News

EleutherAI releases massive AI training dataset of licensed and open domain text
EleutherAI, an AI research organization, has released what it's claiming is one of the largest collections of licensed and ...
Q&A: Leveraging the power of open source data to map the world's oceans
June 8 was UN World Oceans Day, and this year's theme is "Wonder: Sustaining What Sustains Us." One researcher who has spent ...
Eleuther AI releases 8TB collection of licensed and open training data
AI research organization Eleuther AI has launched a massive text database, Common Pile v0.1, that can be used to train AI ...
Tech Xplore on MSN3h
Vision-language models gain spatial reasoning skills through artificial worlds and 3D scene descriptions
Vision-language models (VLMs) are advanced computational techniques designed to process both images and written texts, making ...
Slator1d
Microsoft Introduces Phi-Omni-ST for AI Live Speech Translation
On June 4, 2025, Microsoft released Phi-Omni-ST, an open-source multimodal language model (LM) designed for direct ...
New OpenAI Sora & Google Veo Competitor Focuses on Storytelling With Its Text-to-Video Tool
This text-to-video tool from AI startup Manus turns prompts into full video stories, challenging OpenAI, Google, and Chinese ...
Top AI researchers say language is limiting. Here's the new kind of model they are building instead.
Top AI researchers like Fei-Fei Li and Yann LeCun are developing a "world" model that doesn't rely solely on language.
Slator3d
Images Improve Cultural Accuracy of AI Translations, Research Finds
Researchers from MBZUAI suggest that images help multimodal large language models generate more culturally accurate AI ...
Interesting Engineering on MSN1d
Meta’s new V-JEPA 2 AI helps robots predict and plan using nothing but raw video data
Meta unveils V-JEPA 2, a world model trained on video to help robots predict real-world physics and act with human-like ...