Google's Gemini has arrived. Sort of.

Today's issue is brought to you by our sponsor - Koala. Accelerate Your Content Creation with this (better than ChatGPT) AI Tool!

📰 No-Noise News 📰

🤖 Google's Gemini Pro, the lightweight version of the more powerful Gemini model, has been launched. Gemini Pro is being used to power Google's ChatGPT competitor, Bard, delivering improved reasoning and planning capabilities. Gemini Ultra, the flagship model, will be released ‘at some point’ next year.

+ Google’s blog about Gemini

+ Google’s blog with additional info about Gemini

+ Google has unveiled AlphaCode 2, an improved version of its code-generating AI model powered by Gemini.

📱🖥️ Google has introduced Gemini AI in the Pixel 8 Pro, providing efficient features like on-device audio summarization and enhanced smart replies within messaging apps. These features work offline, ensuring data privacy while boosting the phone's capabilities, potentially influencing users to switch to Android.

🎨 Meta has launched a new standalone AI-powered image generator called “Imagine with Meta”. Similar to OpenAI's DALL-E, the tool allows users to create high-resolution images by describing them in natural language. It's free to use, but Meta plans to add watermarks in the future for increased transparency.

+ Meta's AI characters, based on real-life celebrities, are now available across its U.S. apps, allowing users to chat with them on WhatsApp, Messenger, and Instagram. The AI characters will support Bing Search and have the ability to remember previous conversations, enhancing the experience of interacting with them.

+ Meta AI introduces new features, including 'reimagine' for creative group chats and support for Reels, to enhance user interaction and content creation.

🗣️ Resemble AI has released a new tool called the Deepfake Detection Dashboard, which uses deep learning to differentiate between real and AI-generated audio. The tool can recognize voice-based deepfake audio across media and offers scalability, accuracy, reliability, and voice isolation capabilities.

🧑 TIME’s CEO OF THE YEAR 2023 - Sam Altman

✍️ "Help me write" AI is coming soon to Chrome for desktop. Google is introducing an AI-powered feature called "Help me write" in Chrome for Windows, Mac, and Linux. This feature allows users to save time by generating appropriate text based on a prompt. The AI will also consider the contents of the webpage for more context. Users will have the option to adjust the writing style of the AI. The feature is currently being developed and is expected to be available in Chrome 122 in February 2024.

🖼️ Sydney-based AI art platform Leonardo.AI has raised $31 million in funding. With seven million users and over 700 million images generated, the platform offers collaboration tools, hosting on a private cloud, and access to APIs for enterprise users.  

🧠 Liquid AI, a new MIT spinoff, aims to build general-purpose AI systems powered by liquid neural networks. Liquid neural networks are smaller and require less compute power compared to traditional AI models. They are also more interpretable and have the ability to adapt their parameters for success over time. Liquid AI has raised $37.5 million in funding and plans to commercialize its technology by competing with foundation model companies building generative AI models.

📚 Nerd section 📚

⚒️ RAG vs. Context-Window in GPT-4: accuracy, cost, & latency - RAG (Retrieval Augmented Generation) combined with GPT-4 delivers excellent performance, at a fraction of the cost compared to using a context window.

📜 Boosting RAG-based intelligent document assistants using entity extraction, SQL querying, and agents with Amazon Bedrock. The article explains how to enhance conversational AI assistants by connecting them to internal knowledge bases, using techniques such as retrieval augmented generation (RAG), metadata extraction, SQL querying, and LLM agents.

We love to hear back from you!