AI Daily News - Meta, GPT-4 Mini, Samsung, and more! pt.1

Introduction

Meta's LOM 3 - The 405 Billion Parameter Beast

Meta is set to release LOM 3, an impressive AI model boasting 405 billion parameters on July 23rd. This multimodal marvel has the capability to handle both images and text seamlessly. However, it won't be available in the EU due to uncertainties surrounding GDPR compliance.

Apple’s 4M System on Hugging Face Spaces

Apple is showcasing its 4M system on Hugging Face Spaces. This innovative system can process a variety of data types including text, audio, video, and even 3D data, providing a highly versatile solution for diverse tasks.

OpenAI’s GPT-4 Mini - The Successor to GPT-3.5

OpenAI has unveiled GPT-4 Mini, a new AI model designed to replace GPT-3.5. The GPT-4 Mini supports both text and vision inputs, featuring a substantial 128,000 token context window to enhance its understanding and generation capabilities.

Nvidia and MSTR's MRR Nemo

In a strategic partnership, Nvidia and MSTR have introduced MRR Nemo, a 12 billion parameter model. This AI model focuses on local deployment, driving efficiency and functionality in localized environments.

Samsung’s Sketch to Draw Feature

Samsung has enhanced its Sketch to Draw feature, now allowing users to insert generated objects directly into photos. This feature leverages AI to provide more creative and interactive photo editing experiences.

Amazon's AI Shopping Assistant - Rufus

Amazon has launched Rufus, an AI-driven shopping assistant designed to enhance and streamline the shopping experience for users, making it smarter and more efficient.

Google Unveils Google Vids

Google has introduced Google Vids, a groundbreaking tool for AI-powered video creation within its Workspace suite. This new feature aims to simplify video content creation, making it more accessible and less time-consuming.