Last Week in AI - A Weekly Unwind
From 6-May-2024 to 11-May-2024
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 🧵👇
Amazon Makes it GenAI Assistant Generally Available 🤖
Amazon has made its generative AI assistant, Amazon Q, generally available to businesses and developers. Initially released in November 2023, Amazon Q integrates deeply with enterprise systems, allowing for tasks like chatting, problem-solving, content generation, and actionable insights. For developers, it offers features like code generation, debugging, and upgrading code versions. For businesses, it aids in making data-driven decisions, automating tasks, and creating reports. Additionally, Q’s integration with Amazon QuickSight enhances business intelligence capabilities by enabling natural language-driven data analysis and visualization.
From Simulated Robot Training to Reality with LLMs 🎮
Teaching robots new skills by first training them in a simulated environment and then transferring those skills to the real world has always been tricky. This process usually requires a lot of manual effort to design and fine-tune reward functions and the physics of the simulation. Researchers at Nvidia have introduced “DrEureka,” a system that uses LLMs to automate and speed up this skill transfer process. All that is needed is the physics simulation of the task you want the robot to learn, and it takes care of the rest.
Microsoft’s AI Model to Compete with OpenAI, Google and Anthropic 🚴♂️
Microsoft is reportedly building its first-ever large AI model, and no it’s not with OpenAI for the GPT series. The company is developing MAI-1, an LLM with 500B parameters. The project is being overseen by Mustafa Suleyman who Microsoft had poached from Inflection AI last month. It’s certain by now that Microsoft is decreasing its dependency on OpenAI’s models for powering almost all its AI products, be it Microsoft Copilot or in GitHub.
Devin’s New Features for More Control and Collaboration 🧑💻
Just a couple of months back, the first autonomous AI software engineer Devin was released by Cognitive Labs which created quite a stir in the AI community, and it has already received some major updates. An X user Andrew Gao who has access to Devin posted a thread on these new features, however, Cognitive Labs has not officially announced them. The new features let you take control and guide Dein when it’s stuck or is going wrong, and also boost its automation.
Apple’s New M4 Chip for Powering Next-gen AI Apps 🚀
Apple finally wrapped up its special event to release the new iPad line-up, highlighting its increased focus on AI integration across various features. The new iPad Air now incorporates the M2 chip, previously exclusive to the iPad Pro, enhancing machine learning tasks and speeding up object and text identification in images. Additionally, the iPad Pro features the revolutionary M4 chip, with a Neural Engine capable of 38 trillion operations per second, enabling complex AI applications, such as isolating subjects from video backgrounds with remarkable speed.
GPT2 Chatbot Appears Again! 🤖
Two weeks back, a new chatbot named “gpt2-chatbot” emerged on LMSys’ chatbot arena, spurring excitement among AI enthusiasts with its advanced reasoning capabilities, leading to speculation that OpenAI developed it in a stealth test. However, it disappeared the next day. Last week, two more chatbots, “im-a-good-gpt2-chatbot” and “im-also-a-good-gpt2-chatbot,” appeared, impressing users with their ability to handle various tasks, from coding the Flappy Bird game in one attempt to solving complex physics problems and responding to rubric inquiries where other large language models failed. Currently, these chatbots are only accessible in the “Battle Mode” of the arena, and there are hints, including a tweet from Sam Altman, that they might be OpenAI's creations.
New Connected Apps Feature Coming in ChatGPT ☁️
ChatGPT will soon have the “Connected Apps” feature, integrating with Google Drive and Microsoft One Drive. This will let you attach documents from the cloud directly in ChatGPT, saving the unnecessary hassle of downloading the document first and uploading it again.
OpenAI is holding a special event on May 13 where it is releasing some ChatGPT and GPT-4 updates. We can expect this feature to be released at this event.
Microsoft Copilot gets Auto-complete and Rewrite feature 📝
Microsoft is making it easier to prompt and interact with its Copilot in Microsoft 365 by introducing features like autocomplete, rewrite, and a prompt manager. These features will assist you in crafting precise prompts by offering autocomplete suggestions and transforming basic inputs into detailed queries. Additionally, the new Catch Up chat interface provides insights and recommendations based on recent activities, while Copilot Lab allows users to create, publish, and manage prompts for team-wide use.
AlphaFold 3 for Advancing Drug Development 🧬
Google DeepMind and Isomorphic Labs have unveiled AlphaFold 3, a new AI model that can predict the 3D structure and interactions of various biomolecules, including proteins, DNA, RNA, and ligands. It builds upon the success of AlphaFold 2 to encompass a wider range of molecules and their interactions. AlphaFold 3 offers unprecedented accuracy, providing insights into the intricate mechanisms of life and accelerating drug discovery.
Context Window Scaled from 8K to 10 Million Tokens 🧠
Google had released its series of small opensource models Gemma with a context window of 8K tokens. A team has scaled the context window from 8K tokens to a staggering 10 Million tokens. This window can encompass thousands of documents, research papers, and even 30-50 novels! Despite its large capacity, Gemma 2B is designed to run efficiently, requiring less than 32GB of memory.
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!
🔗 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