Welcome to Get The Gist, where every weekday we share an easy-to-read summary of the latest and greatest developments in AI—news, innovations, and trends—all delivered in under 5 minutes! ⏱

In today’s edition:

• Mistral AI Unveils Ministral 3B and 8B Models for Edge Computing

• Nvidia Quietly Launches AI Model that Outperforms GPT-4

• YouTube Rolls Out AI Music Tool “Dream Tracks” to U.S. Creators

• Google Gemini Can Now Generate Images in Customizable Aspect Ratios

• And more AI news….

Top Developments

Mistral AI

Mistral AI Unveils Ministral 3B and 8B Models for Edge Computing

The Gist: Mistral AI has launched two new AI models, Ministral 3B and 8B, designed for efficient on-device and edge computing. These models outperform competitors and are tailored for tasks requiring privacy-first, local inference.

Key Details:

• Models handle large context lengths (up to 128k) for smooth performance in resource-limited environments.

• Ideal for applications like smart assistants, local analytics, and robotics, enhancing task efficiency.

• Available for commercial use with competitive pricing and research access for the 8B Instruct model.

• Outperforms AI models like Gemma 2 and Llama 3 in benchmarks.

Nvidia

Nvidia Quietly Launches AI Model that Outperforms GPT-4

The Gist: Nvidia has released a new AI model, Llama-3.1-Nemotron-70B-Instruct, which surpasses industry giants like OpenAI's GPT-4 in performance benchmarks. This launch marks a significant expansion of Nvidia’s AI strategy, shifting from hardware to high-performing AI software.

Key Details:

• Nvidia’s new model scored higher than GPT-4 on key benchmarks, demonstrating superior language understanding and generation.

• Developed using advanced techniques like Reinforcement Learning from Human Feedback (RLHF), the model excels in handling complex queries.

• Nvidia offers free access through its platform, allowing businesses to experiment with this powerful AI tool.

• The model is customizable for business needs but requires careful use in specialized areas like legal reasoning or math.

Youtube

YouTube Rolls Out AI Music Tool “Dream Tracks” to U.S. Creators

The Gist: YouTube has launched its AI-powered music generator “Dream Tracks” in the U.S., allowing creators to use text prompts to create custom audio for their short-form videos. The tool aims to deepen the connection between artists and fans through music creation.

Key Details:

• Powered by Google DeepMind’s Lyria, Dream Tracks generates custom instrumental soundtracks for YouTube Shorts.

• U.S. creators can now use this tool to create royalty-free soundtracks up to 30 seconds long.

• Users can remix the AI-generated audio clips, enhancing creative possibilities.

• YouTube applies a hidden SynthID watermark to all AI-generated tracks to ensure transparency.

Quick Gist

Clerk Chat secured $7 million in funding led by Race Capital to enhance its AI-powered business communication platform (Read More).

Anthropic CEO Dario Amodei published a lengthy blog outlining a utopian vision for the transformative potential of artificial general intelligence, while simultaneously seeking to secure a $40 billion valuation for the company (Read More).

Google Cloud announced the general availability of its upgraded Vertex AI platform and Healthcare Data Engine to enhance AI applications in healthcare (Read More).

Amazon led a $500 million funding round for X-energy to roll out 5GW of small nuclear reactors by 2039, while Google partnered with Kairos Power to install 500MW of SMRs by 2035, both aiming to meet rising energy demands from data centers with clean power (Read More).

Google is launching its Gemini AI models for public sector agencies within Google Distributed Cloud in early 2025, along with funding to upskill the government workforce in responsible AI practices (Read More).

Google's Gemini AI chatbot is set to introduce a feature allowing users to generate images in customizable aspect ratios, enhancing its image editing capabilities (Read More).

THAT’S IT FOR TODAY!