It was yet another thrilling week in the AI field with advancements that further extend the limits of what can be achieved with AI.

Here are 10 AI breakthroughs that you can’t afford to miss 🧵👇

Microsoft researchers developed MInference, a new method to address the computational bottleneck in LLM inference with long prompts. It analyzes and leverages patterns in how LLMs pay attention to text, making the whole process much faster without compromising the LLM’s accuracy.

MInference makes pre-fill up to 10x faster for 1 million token prompts on a single A100 GPU. Tasks that previously took half an hour might now only take a few minutes.

At the World Artificial Intelligence Conference, Chinese AI company SenseTime introduced its new multimodal AI model SenseNova 5o and the improved language model SenseNova 5.5. SenseNova 5o can process text, images, audio, and video, and is suitable for real-time interactions. The model outperforms state-of-the-art models like GPT-4o and Claude 3.5 Sonnet across most of the standard benchmarks.

YouTube is updating its “Erase Song” feature with a new AI feature that will identify and remove copyrighted music from videos while keeping other audio elements intact. Creators can choose between “Erase Song” to target only the copyrighted audio or “Mute All Sound” to silence everything within specified timestamps.

Microsoft released AutoGen Studio, a low-code interface to create multi-agent AI applications. This new tool builds on the existing AutoGen framework for developers to build, test, and share AI agents and workflows. With AutoGen Studio, developers can create complex agent interactions with minimal coding. The goal is to make building and deploying these sophisticated AI systems accessible to everyone.

Artifacts that appears in Claude.ai can now be published and shared with others. Not just this, you can take an Artifact shared with you in a new chat with Claude and remix it with your own unique spin!

Samsung wrapped its Unpacked Event 2024 yesterday, revealing the next generations of Z Fold and Flip phones, Galaxy buds, watch, and the highly-anticipated Samsung Ring. Throughout the event, AI took centre stage integrating into every product through Galaxy AI powered by Google’s Gemini AI model.

• Samsung Health app is now powered by Galaxy AI to track vitals and interpret them to give you actionable insights.

• The new AI-powered Galaxy Ring is designed to be worn 24/7 to track your health and vitals throughout day and night.

• New Generative AI features in Galaxy AI include Composer to draft emails and messages, transcription and summarizing audio notes, sketch-to-image, Live Translate in 16 different languages, and more.

Prompt quality is crucial for AI application results. Anthropic Console now offers tools to generate, test, and evaluate prompts within a streamlined workflow, eliminating the need for manual testing.

It leverages Claude 3.5 Sonnet to generate effective prompts and test cases. You can create and manage extensive test suites, compare multiple prompt versions, and even have experts grade response quality on a 5-point scale.

Microsoft has relinquished its observer seat on OpenAI’s board, a position it held for less than eight months. Apple too was initially slated to appoint an observer on OpenAI’s board, but has also opted out. OpenAI will now engage with Microsoft and Apple through regular stakeholder meetings.

AWS has introduced AWS App Studio, currently in public preview, for users to build enterprise-grade custom applications quickly, using natural language instead of traditional coding.

Just describe your desired application in plain language, and App Studio’s AI will generate the core components, including the UI, data structures, and basic logic. It also has pre-built connectors for services like Amazon Aurora, DynamoDB, S3, and Salesforce to integrate data for applications.

LMSYS has developed RouteLLM, an open-source framework for cost-effective LLM routing. This system acts as an intelligent gatekeeper, analyzing incoming queries and intelligently directing them to the most appropriate LLM based on both the task’s complexity and the capabilities of available models.

RouteLLM achieved cost savings of over 85% on the MT Bench, 45% on MMLU, and 35% on GSM8K compared to using only GPT-4, all while maintaining 95% of GPT-4’s performance.

Which of the above AI development you are most excited about and why?
Tell us in the comments below ⬇️

That’s all for today 👋

Stay tuned for another week of innovation and discovery as AI continues to evolve at a staggering pace. Don’t miss out on the developments – join us next week for more insights into the AI revolution!

Click on the subscribe button and be part of the future, today!

📣 Spread the Word: Think your friends and colleagues should be in the know? Click the ‘Share’ button and let them join this exciting adventure into the world of AI. Sharing knowledge is the first step towards innovation!

Share

🔗 Stay Connected: Follow us for AI updates, sneak peeks, and more. Your journey into the future of AI starts here!

Shubham Saboo – Twitter | LinkedIn ⎸ Unwind AI – Twitter | LinkedIn

Awesome LLM Apps | Sponsor Us