An Unusual week in AI...

This week's newsletter explores three groundbreaking AI developments: OpenAI's SearchGPT, a prototype for conversational web searches with clear sources; Mistral AI's Mistral Large 2, a powerful multilingual language model with a 128k context window; and Kling AI's global launch of an AI video generator creating videos from text prompts. These advancements showcase rapid progress in search technology, language processing, and content creation, potentially transforming our interaction with information and content creation. Let's dive deeper into how these innovations could shape AI's future.

Stuff you should know

OpenAI Unveils SearchGPT Prototype

OpenAI is currently testing SearchGPT, a prototype designed to enhance search capabilities by providing fast and relevant answers sourced from the web. This new feature aims to streamline the search process by integrating AI's conversational strengths with real-time information, allowing users to receive direct responses to their inquiries along with clear citations to relevant sources. The prototype is being launched to a select group of users and publishers for feedback, with the intention of refining the experience before integrating successful features into ChatGPT. SearchGPT allows users to engage in a conversational manner, asking follow-up questions while maintaining context. OpenAI emphasizes collaboration with publishers to ensure that high-quality content is highlighted and easily accessible, fostering a more interactive search environment. The company is also providing publishers with tools to manage their visibility in SearchGPT, ensuring they can engage effectively with the AI-driven search landscape. Feedback from this testing phase will guide future improvements, particularly in areas like local information and commerce

Mistral AI Launches Mistral Large 2 Model

Mistral AI has launched Mistral Large 2, a state-of-the-art language model featuring a 128k context window and 123 billion parameters, designed for high-performance single-node inference. This model supports multiple languages, including major European and Asian languages, as well as over 80 coding languages. Mistral Large 2 excels in cost efficiency and performance, achieving an accuracy of 84.0% on the MMLU benchmark, and competes with leading models like GPT-4o and Claude 3 Opus. Significant improvements have been made in reasoning capabilities, reducing the model's tendency to generate inaccurate information, and enhancing its ability to follow instructions in multi-turn conversations. Mistral Large 2 is now available for use on la Plateforme and can be accessed through major cloud providers, including Google Cloud and Azure, allowing for broader application in AI-driven tasks and multilingual document processing.

Kling AI Launches Global AI Video Generator

Kling AI has made its AI-powered video generator available worldwide, allowing users to create videos by inputting text prompts. The platform features a user-friendly interface, advanced AI technology, and the ability to generate high-definition videos up to two minutes long, supporting both text-to-video and image-plus-text-to-video generation. With 66 free daily credits, users can experiment with video creation without immediate costs. Kling AI aims to democratize video production, making it accessible to individuals without prior video editing experience. While the system has limitations, such as restrictions on politically sensitive content and slower generation times, it positions itself as a promising tool for content creators, businesses, and marketers looking to leverage AI in their video production workflows.

Hand Picked Video

In this video, we'll explore the pros and cons of open source AI Models. We'll examine the advantages and drawbacks of using and contributing to open source projects, providing a balanced view of this important aspect of the software development ecosystem.

Top AI Products from this week 

• Brev.ai - Brev.ai is an AI music generator that turns text into music using AI technology. Create high-quality songs and instrumentals effortlessly, perfect for videos, social media, and more. Just enter your text, generate, and download your music.

• PixVerse V2 - Exciting News! PixVerse just got an upgrade! Introducing PixVerse V2🔥 • 8-second video generation • Enhanced details and motion • Consistent style, subject, and scenes across 1-5 clips for seamless storytelling

• Kotae - Automate 80% of Customer Inquiries in 80+ Languages with Kotae – Your AI Chatbot Powered by Your Content. Ready in 5 Minutes, No Coding Required. Enjoy a Free 30-Day Trial with Personalized Onboarding.

• Last24.ai - Last24 is an AI search engine that helps you understand today's news fast. It searches the internet, picks the important news you need, and summarizes the key points in a beautiful mindmap.

• Team-Up - Connect with startup enthusiasts through spontaneous live interactions. Find mentors, talents, and partners. Share ideas, seek advice. Expand your network, uncover opportunities. Secure platform for spontaneous business networking and unexpected connections.

• Topview.ai - Topview is an online AI video editor that turns your links or media assets into viral videos in one click, powered by Youtube, Tiktok, & Facebook ads library.

• Datrics AI Analyst Builder - AI Analyst Builder enables teams to create custom AI analysts without coding. These analysts answer data questions via a chat interface like ChatGPT. Tailored to specific business processes and data, it ensures accurate results and avoids hallucination.

This week in AI

• Udio v1.5 Enhances AI Music Generation - Udio has released version 1.5 of its AI music generator, improving audio quality and user experience. The update features human-like vocals, audio-to-audio remixing, and enhanced language support, narrowing the gap between AI and human-created music

• Microsoft Introduces Serverless Fine-Tuning for Phi-3 Model - Microsoft has unveiled serverless fine-tuning for its Phi-3 small language model, enabling developers to adapt the model without managing servers. Phi-3, a 3 billion parameter model, offers affordable performance for coding and reasoning tasks. The new service is available on Azure AI, allowing quick access for developers. This update positions Microsoft as a strong competitor in the AI space, especially against OpenAI's offerings.

• Free Gemini Users Can Now Chat Instantly - Google has made Gemini 1.5 Flash available to free users, allowing them to engage in quick conversations with the chatbot. This update also includes related links in responses, enhancing the overall user experience and accessibility of the AI tool.

• Enhance Your Videos with AI-Powered Eye Contact - The Rundown tutorial explores using AI-powered eye contact to make videos more engaging. This technology simulates direct eye contact, improving viewer connection and retention. Practical tips are provided for integrating this feature into video projects.