Today’s top AI Highlights:

1. GitHub’s new Playground brings the best AI Models for free within GitHub

2. Cloudflare’s new AI Playground to chain multimodal AI models

3. Opensource image generation models beat DALL.E-3 and Midjourney

4. Google’s new experimental Gemini 1.5 Pro available to use

5. Build AI apps on top of leading AI models by simply telling

& so much more!

Read time: 3 mins

Latest Developments

AI Models Integrated Where You Code 🧑‍💻

GitHub has just launched GitHub Models in limited public beta, a built-in playground that lets you test leading AI models for free right within GitHub. From Llama 3.1 to GPT-4o and GPT-4o mini, to Phi 3 or Mistral Large 2, you can access each model in this playground and test different prompts and model parameters, before moving to a familiar coding environment in Codespaces and VS Code, and finally deploying with Azure AI for production-level support.

Key Highlights:

1. Interactive Model Playground - This free playground provides a sandbox environment to test various AI models with different prompts and parameters so you can quickly compare and experiment without leaving the GitHub platform.

2. Seamless Codespaces Integration - Once you've experimented in the playground, GitHub Models makes it easy to transition into development within Codespaces. Pre-built code samples in multiple languages and frameworks are available, streamlining prototyping and building proof-of-concept applications.

3. Simplified Production Pathway - GitHub Models offers a smooth transition to Azure AI for production deployments, giving you a reliable and secure platform for running your AI applications.

Chain Multiple Multimodal AI Models Together ⛓️

Cloudflare has released a multimodal AI playground to experiment with various AI models and chain them together to build interesting AI applications. The playground is designed to be user-friendly with a drag-and-drop interface. You can explore and build AI workflows for various real-world applications like content generators, image captioning tools, etc.

Key Highlights:

1. Model diversity - The playground includes a variety of multi-modal AI models like Llama 3.1, Stable Diffusion, Llava 1.5, which can process text, images, and even audio.

2. Flexibility - The node-based interface lets you connect different models together in various configurations. You can create unique workflows and explore how different models can complement and enhance each other.

3. Try it Out - The playground comes with pre-loaded example workflows to give you a starting point for your experiment.

Opensource Contender to Midjourney and DALL.E-3 🏆

After Stable Diffusion, here’s a new strong contender in the opensource arena for text-to-image models. Black Forest Labs has just launched, along with their opensource image generation suite of models called FLUX.1. This suite of models delivers state-of-the-art performance in image detail, prompt adherence, style diversity, and scene complexity, outperforming even DALL.E-3 and Midjourney v6.

Key Highlights:

1. Three Models - FLUX.1 is available in three versions: [pro] for top-tier performance via API access, [dev] an open-weight version for non-commercial use, and [schnell] for local development and personal use under an Apache 2.0 license. You can try [pro] and [dev] now.

2. Transformer Models - All public FLUX.1 models are based on a hybrid architecture of multimodal and parallel diffusion transformer blocks and scaled to 12B parameters.

3. Performance - FLUX.1 models surpass other leading models like Midjourney’s latest v6.0, OpenAI’s DALL·E 3, and Stable Diffusion 3-Ultra in various aspects, including visual quality, prompt following, and output diversity. All FLUX.1 models support various aspect ratios and resolutions.

Quick Bites

1. Google has made an experimental version of Gemini 1.5 Pro (0801) available for early testing in Google AI Studio and the Gemini API. This version of Gemini 1.5 Pro is currently #1 on the LMSYS Chatbot Arena Leaderboard for both text and multi-modal.

2. The EU’s AI Act has officially come into force as of August 1, 2024, starting a series of staggered compliance deadlines for AI developers and applications. High-risk AI users will need to meet specific risk and quality management obligations, while low-risk and general-purpose AI will have lighter transparency requirements.
   

3. OpenAI CEO Sam Altman announced that OpenAI will provide early access to its next major generative AI model to the U.S. AI Safety Institute for safety testing. He says they are “committed to allocating at least 20% of the computing resources to safety efforts.”
   

4. Google is adding Gemini-powered features to Chrome desktop, including Google Lens in Chroma, tab comparison for shopping, and natural language search history. These updates will let you identify items, compare products across tabs, and search your browsing history using simple questions, rolling out in the coming weeks.

Tools of the Trade

1. WorkflowAI: Build AI apps on top of leading LLMs using natural language. It allows seamless model comparison and switching, easy integration into your codebase, and cost-effective deployment without compromising quality.

2. Martin: Your personal AI butler that you can text, email, or talk to. It can handle tasks like scheduling, briefings, and communications. Integrated with your calendar, inbox, and contacts, Martin proactively assists you and understands your preferences and needs over time.

3. CongressGPT: Understand what Congress is actually doing. Chat with real bills from the U.S. House and Senate, with over 199 bills available and new ones added nightly.

4. Awesome LLM Apps: Build awesome LLM apps using RAG for interacting with data sources like GitHub, Gmail, PDFs, and YouTube videos through simple texts. These apps will let you retrieve information, engage in chat, and extract insights directly from content on these platforms.

Hot Takes

1. Ultimately, you can only learn to function in the real world by interacting with it. Therefore LLMs cannot solve AI. ~
   Pedro Domingos
   

2. I still think OpenAI will release something as open source this year.
   4o-mini would have been a perfect candidate and there's not much risk in releasing it for download, but let's wait and see. ~
   Flowers

Meme of the Day

That’s all for today! See you tomorrow with more such AI-filled content.

Real-time AI Updates 🚨

⚡️ Follow me on Twitter @Saboo_Shubham for lightning-fast AI updates and never miss what’s trending!

PS: We curate this AI newsletter every day for FREE, your support is what keeps me going. If you find value in what you read, share it with your friends by clicking the share button below!