Apple and Google Unveil New AI Developments:
MLX Framework and NotebookLM App

In a week filled with groundbreaking AI announcements, both Apple and Google have taken significant strides in advancing their capabilities in the artificial intelligence landscape. A Revolutionary Step in Machine Learning for Apple Devices. Amidst the buzz surrounding Google’s unveiling of the multimodal model, Gemini, Apple made a quieter yet impactful move by introducing its AI framework, MLX. Designed specifically for Apple Silicon chips, MLX aims to pave the way for efficient machine learning on devices such as MacBooks. Unlike Google’s Gemini, which is a large language model (LLM), MLX focuses on creating a framework optimized for Apple Silicon. The framework’s design draws inspiration from well-established frameworks like PyTorch, Jax, and ArrayFire.

What sets MLX apart is its unified memory model, allowing operations on MLX arrays without the need for data copies, supporting both CPU and GPU. Awni Hannun, a machine learning researcher at Apple, highlighted MLX’s efficiency, describing it as “an efficient machine learning framework specifically designed for Apple silicon.” While this development may be more pertinent for Apple developers currently, its significance will become apparent to end-users as generative AI tools and apps start making their way to MacBooks. The announcement aligns with previous reports hinting at Apple’s interest in incorporating generative AI features into its products, evident from recent job listings related to AI.

On the other side of the spectrum, Google has officially launched NotebookLM, an AI-powered note-taking app enriched with the capabilities of its Gemini Pro large language model (LLM). This experimental app, formerly known as Project Tailwind, was initially revealed at Google I/O 2023 and has undergone substantial development. NotebookLM, available for users in the US aged 18 and above, promises to transform the note-taking experience. Powered by Gemini Pro, the app offers features that seamlessly integrate reading, note-taking, and writing.

Highlighted Features of Google’s NotebookLM:

1. Dedicated Noteboard Space: Users can create a centralized hub for organized note-taking, pinning quotes, excerpts, and personal notes.
2. Citation Integration: Automatic citation sharing enhances context, allowing users to navigate from a citation to the source effortlessly.
3. Save Responses with Citations: Users can save responses from NotebookLM as notes, complete with citations.
4. Suggested Actions: The platform suggests actions based on user activities, aiding in summarizing text and understanding technical language.
5. Tools for Note Refinement: NotebookLM offers tools to refine writing, assisting users in creating well-crafted notes and suggesting related ideas based on recent writing.
6. Structured Document Creation: Users can organize curated notes into structured documents, with the platform suggesting different formats such as thematic outlines or study guides.
7. Export to Google Docs: A one-click export feature allows users to seamlessly transition to Google Docs for final document editing.

While NotebookLM is currently accessible to select users in the US, Google emphasizes that it is an ongoing experiment, poised to evolve based on user feedback. In a week filled with AI revelations, Apple and Google have both positioned themselves at the forefront of innovation, bringing forth advancements that will shape the future landscape of artificial intelligence applications and user experiences.