Schools nationwide abandon unreliable AI detection software that falsely accused students of cheating and adopt transparency-based policies requiring students to disclose AI usage while redesigning assignments to promote genuine engagement and learning.
Google's AI assistant Gemini will now automatically summarize long emails, displaying key points at the top, unless users opt out of this feature, marking AI's increasing presence in everyday software and services despite concerns over accuracy.
New Transformer models employ techniques like sparse attention, low-rank factorization, and kernel approximations to overcome quadratic complexity, enabling efficient processing of longer sequences while preserving long-range dependency modeling capability.
Chinese AI startup DeepSeek quietly released an upgraded version of its reasoning model DeepSeek R1, outperforming rivals like Meta and OpenAI, shocking markets and sparking concerns about U.S. tech giants' spending, while showcasing China's AI capabilities despite U.S. export curbs.
A revolutionary framework unleashes the power of over 100 AI models, enabling cutting-edge training with efficient fine-tuning, advanced algorithms, and diverse approaches for language and vision tasks.
In a groundbreaking deal, Elon Musk's AI company xAI will pay Telegram $300 million to integrate its chatbot Grok into the messaging app, allowing users to access Grok's advanced capabilities for writing, summarizing, and creating content directly within Telegram.
Meta faces an exodus of top AI talent, including key researchers behind its open-source Llama models, as growing competition from rivals like Mistral lures away employees with lucrative offers, raising concerns about Meta's ability to maintain its lead in the AI race.
Anthropic rolls out a beta voice mode for its Claude chatbot apps, allowing users to have spoken conversations with the AI assistant and enabling hands-free interactions while displaying key points on-screen and offering voice response options.
Sarvam AI's latest release of Sarvam-M, a 24B open-weights hybrid language model for Indian languages, has sparked debate on the state of Indic AI research, with critics questioning its impact and funding, while supporters highlight its importance for solving Bharat-centric use cases.
AI models can now translate natural language into SQL, enabling non-technical users to access data efficiently; techniques like context building and intent understanding improve translation quality, evaluated through synthetic benchmarks and AI-based judging.