Gen AI Updates: Mistral Large's Arrival, Alibaba's EMO, Data Moves by OpenAI & MidJourney, Adobe's AI Music & Amazon's Privacy Measures

Welcome to the eighth edition of our weekly newsletter ??, your go-to source for the latest developments and trends in Generative AI.

Each edition brings you a curated selection of impactful news, insightful analyses, and exciting advancements from the dynamic world of generative AI. Stay tuned for a concise and informative exploration of this rapidly evolving field.

1. Mistral AI Unveils Mistral Large

Mistral AI announced its flagship model, Mistral Large, showcasing top-tier reasoning abilities and available via Azure. As the second-ranked globally available model, it excels in multilingual tasks, code generation, and complex text transformations. Featuring native fluency in multiple languages, a vast context window, and precise instruction-following capabilities,

Mistral Large also introduces function calling for scalable application development. Alongside, Mistral Small is launched for latency-sensitive applications, both aiming to make advanced AI accessible and efficient for developers. Read more

2. Alibaba's EMO: Bringing Photos to Life with AI

Alibaba's new AI framework, EMO, transforms any photo into a talking or singing video using just a single image and audio input. This groundbreaking framework outperforms traditional methods by capturing intricate facial expressions and movements, creating lifelike videos with expressive detail.

EMO's deep neural network, powered by diffusion models, excels in generating videos that maintain the subject's identity, offering applications in entertainment, telepresence, and more. With its ability to animate various portrait styles, EMO sets a new standard for realism and expressiveness in digital content creation. Read more

3. OpenAI & MidJourney Eye WordPress and Tumblr Data

In a bold move signaling the escalating demand for AI training data, OpenAI and MidJourney are in discussions with Automattic, the owner of WordPress and Tumblr, to acquire user-generated content from these platforms. This development underscores a shift towards securing paid data sources amid growing copyright concerns. Automattic is considering introducing an opt-out feature for users unwilling to share their data, albeit with potential visibility penalties for their accounts. This situation highlights the intensifying debate over data privacy and the ethical use of user-generated content in AI development. Read more

4. Adobe Unveils AI Music Creation Tool

Adobe has revealed Project Music GenAI Control, an innovative tool that generates and customizes music from text or melodies. Still in research with no UI, it promises creative freedom without the need for composing skills.

Amidst AI's ethical debates, Adobe prioritises licensed or public domain data for development, also exploring watermarking for produced audio identification. Read more

5. Amazon Sets Boundaries on AI Tool Use

Following the lead of Apple and Walmart, Amazon has imposed restrictions on its employees' use of generative AI tools like ChatGPT for work, aiming to protect sensitive company data and mitigate third-party content claims. This trend among tech giants underscores the critical need for enhanced PII reduction and rigorous data privacy practices to securely manage AI applications in the workplace, amidst intensifying ethical scrutiny. Read more

?? Foundation Model of the Week - Vicuna 13B

Vicuna-13B is an open-source chatbot based on LLaMA-13B. It was developed by training LLaMA-13B on user-shared conversations collected from?ShareGPT. LLaMA is an?open-source language model from Meta Research?that performs as well as comparable closed-source models. Using GPT-4 to evaluate model outputs, the developers of Vicuna-13B found that it not only outperforms comparable models like Stanford Alpaca but also reaches 90% of the quality of OpenAI's ChatGPT and Google Bard.

Try it on Katonic Playground: vicuna-13b

Subscribe for more exciting AI updates in the future. Have a great weekend! ?