Monitor new model and product release from leading AI companies on X (Twitter).
New model or AI product release
OpenAI announces the Stargate Project, a $500 billion initiative to enhance AI infrastructure in the U.S., aiming to secure leadership in AI, create jobs, and bolster national security.
SnowflakeDB AI Research has launched SwiftKV, optimizing Llama models to reduce inference costs by up to 75%. These models, based on Meta's Llama 3.3 and 3.1, are now available on Cortex AI.
Microsoft announces the launch of Microsoft 365 Copilot Chat, aimed at making AI tools more accessible for employees across various businesses, enhancing productivity and collaboration.
Meta has launched the Contextual AI Platform, an enterprise solution for creating specialized RAG agents that enhance expert knowledge work, achieving state-of-the-art accuracy from the start.
OpenAI has launched a beta version of 'tasks' for ChatGPT, allowing users to schedule one-time or recurring actions, such as reminders or personalized briefings, enhancing user interaction.
OpenAI has introduced a new feature allowing ChatGPT users to schedule tasks, such as weekly news briefings or personalized workouts, currently available in beta for Plus, Pro, and Teams users.
OpenAI has launched a beta version of 'tasks' for ChatGPT, allowing users to set one-time reminders or recurring actions, enhancing the assistant's ability to manage future requests automatically.
xAI has launched a new standalone Grok iOS app, enabling users to harness powerful AI for generating images and accessing personalized real-time news and sports updates.
NVIDIA's Project DIGITS will allow developers to prototype and fine-tune Llama models locally, promoting innovation. Ahmad Al Dahle emphasizes the commitment to democratizing AI technology through open-source access.
NVIDIA has launched the Llama Nemotron and Cosmos Nemotron model families, designed to develop advanced AI agents capable of communication, vision, and action, enhancing enterprise agentic AI capabilities.
Meta FAIR has introduced the Byte Latent Transformer (BLT), which matches the performance of tokenization-based LLMs while enhancing inference efficiency and robustness, marking a significant advancement in AI technology.
SemiKong, the first open-source semiconductor-focused language model built with Llama, empowers semiconductor companies to create Domain-Expert Agents, enhancing their ability to leverage specialized knowledge.