Note from Matt: Below is an AI-generated list of AI news and tools. Our bot scours the internet daily to find AI news. Because it’s AI, it often finds older news or non-AI news and adds it to the mix. It also seems to break links often. We’re constantly working on improving the automation but this should, for the most part, ensure no AI news ever gets missed

The Latest AI News:

• Meta releases Llama 3.1 – Meta launched its new flagship open-source model series, Llama 3.1, featuring a 405B parameter version, a 128K context window, and enhanced coding and vision capabilities.
• OpenAI plans new model to counter Llama 3.1 – OpenAI is reportedly preparing to release a new, more powerful and efficient model soon to compete with Meta’s newly launched Llama 3.1.
• Apple in talks with Google, Anthropic for AI chatbots – Apple is reportedly negotiating with Google, Anthropic, and other AI companies to integrate their chatbots into Apple Intelligence for handling more complex queries.
• Claude 3.5 Sonnet now available on Amazon Bedrock – Anthropic’s fastest and most intelligent model, Claude 3.5 Sonnet, is now generally available for building generative AI applications via Amazon Bedrock.
• EU regulators to launch full-scale probe into Microsoft’s OpenAI investment – European Union antitrust regulators are reportedly set to launch a full-scale investigation into Microsoft’s multi-billion dollar investment in OpenAI over competition concerns.
• Humane is looking for a buyer, with HP as a possibility – Humane, the startup behind the Ai Pin, is reportedly seeking a buyer for its business, with HP emerging as a potential suitor.
• DOJ and FTC ramp up antitrust investigations into AI giants – The Justice Department and Federal Trade Commission are advancing their antitrust probes into Microsoft, Nvidia, and OpenAI, requesting extensive information on their partnerships and market conduct.
• Leaked document reveals high costs of Google’s AI search – A leaked internal document shows that expanding Google’s Search Generative Experience globally could more than double its multi-billion dollar operational costs.
• Character.AI in talks for new funding at $5B valuation – AI chatbot startup Character.AI is reportedly in discussions to raise new capital that would value the company at $5 billion.
• Reka launches Reka Core, its flagship multimodal model – Reka AI has released Reka Core, a highly capable multimodal model trained on text, images, audio, and video that shows competitive performance against GPT-4o and Gemini 1.5 Pro.
• ElevenLabs launches AI sound effects generator – Voice AI company ElevenLabs has released a new tool that can create realistic sound effects and short instrumental tracks from text prompts.
• Perplexity launches ‘Convert to Post’ feature – The conversational search engine introduced a new feature that instantly transforms any search thread into a formatted blog post, tweet, or LinkedIn article.
• Microsoft is looking to build its own in-house AI server chips – A new job posting suggests Microsoft is developing its own AI server chips, aiming to reduce its reliance on Nvidia’s GPUs.
• MindEye2 can reconstruct high-resolution images from brain activity – Researchers have developed MindEye2, a new model that can reconstruct complex, high-quality images from fMRI brain recordings with remarkable accuracy.
• US Army deploys Palantir AI to wage information warfare – The US Army is utilizing a Palantir-developed AI tool to analyze online data from platforms like X and Telegram to counter foreign disinformation campaigns.
• Elon Musk gives updates on Tesla’s FSD and AI Day 3 – Elon Musk noted major improvements in FSD v12.4 and confirmed that Tesla’s AI Day 3 event will take place on September 19th.
• China surpasses the U.S. in AI patents – A new report shows that China has overtaken the United States in the number of AI-related patents filed, highlighting its focus on dominating the technology.
• AI PCs vs. traditional PCs: What’s the difference? – A detailed comparison explains the key differences between new AI PCs with NPUs and traditional computers, and whether the upgrade is currently worth it.
• Voice actor protection act could change how games use AI voices – A new US bill aims to protect voice actors from unauthorized AI replication of their voices, which could have significant implications for the video game industry.
• SEC adopts rules to address AI-related conflicts of interest – The U.S. Securities and Exchange Commission has implemented new rules requiring investment firms to manage conflicts of interest arising from their use of AI in investor interactions.
• Claude 3.5 Sonnet now available on Google Cloud’s Vertex AI – Following its launch on AWS, Anthropic’s Claude 3.5 Sonnet model is now also available to customers through Google Cloud’s Vertex AI platform.
• Mistral AI releases Codestral 22B – Mistral AI launched Codestral 22B, a new powerful and fast open-weight model specifically designed for code generation tasks.
• Adept releases Fuyu-Heavy model that beats GPT-4o on web tasks – Adept AI has launched Fuyu-Heavy, a new multimodal model built for digital agents that outperforms GPT-4o on benchmarks for interacting with websites and applications.
• Google’s NotebookLM gets Gemini 1.5 Pro and student features – Google updated its AI-powered research tool NotebookLM with the more powerful Gemini 1.5 Pro and added new features to help students create study guides and prep for interviews.
• Leonardo AI releases Phoenix image generation model – Leonardo AI has launched Phoenix, a new generative model praised for its speed, coherence, and superior ability to follow complex text prompts for image creation.
• How to fine-tune Meta-Llama-3.1-8B-Instruct with Hugging Face – A technical guide provides the steps and code required to fine-tune Meta’s new Llama 3.1 8B model using Hugging Face’s TRL library.
• How to Build a GPT-4o-like Voice AI with Python – A GitHub repository provides a comprehensive guide and pipeline for building a conversational voice AI assistant similar to GPT-4o using various APIs.
• Run Llama-3.1-405B on any cloud with a single command – The SkyPilot framework has been updated to allow developers to easily deploy and run Meta’s largest Llama 3.1 model on any major cloud provider.
• Stability AI’s Stable Diffusion 3 Medium is now available on Amazon Bedrock – The latest and most advanced text-to-image model from Stability AI can now be accessed by developers through Amazon’s managed AI service.
• The Browser Company pivots to building an AI-native computer – The company behind the Arc browser announced it is shifting its focus to create a new type of “AI-native” computer that can perform tasks on behalf of the user.
• Contextual.io raises $20M to automate complex enterprise work – A new startup co-founded by former Google AI leaders has raised $20 million to build AI agents capable of automating complex, multi-step business workflows.
• Together AI and Databricks integration for fine-tuning models – Together AI announced a partnership with Databricks, enabling customers to securely connect private data from the Databricks platform to fine-tune open-source models.
• GitHub Copilot Enterprise in Visual Studio Code is now generally available – The enterprise version of GitHub Copilot, which can be customized to an organization’s private codebase, is now fully available as a VS Code extension.
• Amazon launches generative AI hub for sellers – Amazon has introduced a new hub for third-party sellers that consolidates generative AI tools for creating product titles, descriptions, and lifestyle images.
• Viggle adds a new ‘/animate’ feature driven by text prompts – The AI video generation tool Viggle has released a feature allowing users to animate a character by simply describing the desired motion in a text prompt.
• Google open-sources new tools to build AI responsibly – Google has released the Responsible AI Toolkit, including the new Fairness and Safety evaluation tools, to help developers assess their models.
• Microsoft’s Phi-3-vision is now available on Hugging Face – Microsoft’s lightweight yet powerful multimodal model, Phi-3-vision, which can analyze both images and text, has been released on the Hugging Face platform.
• An interview with Perplexity CEO Aravind Srinivas – Perplexity’s CEO discusses the company’s strategy, the role of ads, its relationship with publishers, and its goal of challenging Google Search.
• Vercel AI SDK 3.2 adds support for OpenAI Assistants v2 and more – The latest version of Vercel’s AI SDK introduces first-class support for the OpenAI Assistants v2 API, client-side tool calling, and other new features.
• What’s new in Llama 3.1 – A technical article explores the new features and architectural changes in Meta’s Llama 3.1, including its vision capabilities and context window.
• Demucs by Meta – An open-source tool that uses AI to separate music tracks into their constituent parts, such as vocals, drums, and bass.
• Build your personal AI data analyst with Llama 3 – A tutorial demonstrates how to create a personal data analyst using Llama 3, the Ollama server, and Streamlit for a user interface.
• Why is AI so good at translating? – An article explains that the success of AI in translation is due to its architecture being well-suited for the task and the vast amount of high-quality training data available.
• Gen-2 by Runway gets major updates – Runway’s text-to-video model Gen-2 has been updated to produce higher-fidelity and more consistent results, along with improved camera controls.
• Udio – A web-based tool that allows users to generate high-quality, full-length music tracks in various genres from simple text prompts.
• A practical guide to fine-tuning LLMs – An in-depth guide covers the essential considerations and best practices for successfully fine-tuning large language models for specific tasks.
• A Practical Guide to LLM Red Teaming – This article provides a comprehensive overview of how to conduct red teaming for LLMs to identify and mitigate potential harms, biases, and security risks.
• The New Future of Work is a Chatbot – An essay argues that conversational AI is becoming the new “operating system” for enterprise software, simplifying complex workflows into a single chat interface.
• Opinion: Llama 3.1 is Overrated – An opinion piece argues that despite impressive benchmark scores, Llama 3.1 does not represent a significant leap forward in reasoning or capability over previous models.
• How custom AI models can help your business – A look at how businesses can gain a competitive edge by developing or fine-tuning custom AI models tailored to their specific data and needs.
• The AI safety problem is a social problem – An article posits that the challenges of AI safety are less about technical alignment and more about addressing complex social, political, and governance issues.
• Google AI presents Simpler One Pass – A research paper from Google introduces an efficient, single-pass decoding algorithm for language models that reduces latency without sacrificing quality.

Trending AI Tools:

• Devin – An autonomous AI software engineer that can handle entire development projects from a single prompt.
• Figure 01 – A humanoid robot powered by OpenAI models that can have full conversations and perform physical tasks.
• Suno – An AI model that generates realistic music, including vocals and lyrics, from a simple text prompt.
• Viggle – A character-to-video AI model that animates a static character image using a motion video as a guide.
• Claude 3 – A family of powerful AI models from Anthropic that excel at a wide range of cognitive tasks.
• Grok – A conversational AI from xAI, designed with a rebellious streak and access to real-time information, which is being open-sourced.
• ChatGPT – OpenAI’s flagship conversational AI used for a wide range of tasks including writing, coding, and brainstorming.
• DeepMind SIMA – A generalist AI agent that can follow natural-language instructions to perform tasks across a variety of 3D virtual gaming environments.
• Krea AI – An AI tool that can generate and enhance images and videos in real-time based on user input and prompts.
• Pika – An idea-to-video platform that allows you to generate and edit videos using AI from text or image prompts.
• Runway – A comprehensive AI-powered video editing and generation suite for creators.
• Midjourney – A popular AI image generator known for producing high-quality, artistic images from text prompts.
• Perplexity – An AI-powered conversational search engine that provides direct answers to questions with cited sources.
• ElevenLabs – An AI voice generation and cloning platform for creating realistic and expressive speech in multiple languages.
• Leonardo AI – An AI image generation platform that recently launched its ‘Phoenix’ model for creating hyper-realistic photos.
• Ideogram – An AI image generator particularly skilled at creating images that include coherent and readable text.
• Google Gemini – Google’s family of powerful, multimodal AI models that can understand and operate across text, images, video, and audio.
• Microsoft Copilot – Microsoft’s AI assistant integrated across Windows, Bing, and Office 365 apps to enhance productivity.
• Stable Diffusion – A foundational open-source text-to-image model that allows users to create detailed images from text descriptions.
• Luma AI’s Dream Machine – A highly anticipated AI model for creating high-quality, realistic videos from text and images.
• Character.AI – A platform that lets you create and interact with AI chatbots based on fictional characters or historical figures.
• Magnific – An AI-powered tool for upscaling and enhancing the details of images to an incredible level.
• Vercel v0 – A generative UI system by Vercel that creates React components and UI layouts based on simple text prompts.
• Coze – A no-code platform for quickly building and deploying your own AI chatbots on various social media and messaging apps.
• Explainpaper – An AI tool that helps you read and understand complex academic research papers by providing simple explanations.
• Gamma – An AI-powered tool that quickly generates polished presentations, documents, and webpages from your notes.
• Co-Star – An AI programming assistant that functions as a junior developer to help fix code, generate unit tests, and explain code blocks.
• Phind – An AI-powered search engine designed specifically for developers to answer technical questions with code examples.
• Magicpatterns – An AI-powered tool for generating UI components, patterns, and even entire screens from text prompts.
• Chatbase – A platform for building a custom ChatGPT-like chatbot for your website using your own documents and data.
• Notion AI – A suite of AI features integrated into the Notion workspace to assist with writing, summarizing, and brainstorming.
• Mindgrasp – An AI learning assistant that can instantly create accurate notes and answer questions from any document, video, or recording.
• Poe by Quora – A platform that provides access to a variety of different AI chatbots and language models all in one place.
• Lindy – An AI assistant that helps automate tasks like email drafting, calendar management, and screening potential clients.
• Context – An AI tool that creates a chatbot from your existing documentation to provide instant answers to user questions.
• Giga ML – A platform for efficiently running and fine-tuning open-source large language models with less computational overhead.
• Udemy – An online learning platform that has integrated new AI-powered features to personalize the learning experience.
• AlphaCode 2 – Google DeepMind’s AI system that excels at solving complex competitive programming problems.

Sponsors:

• Braintrust – A network for senior tech talent to find AI project work with top-tier companies.
• beehiiv – A newsletter platform built for growth and used by some of the largest newsletters in the world.
• RunPod – A platform to rent GPUs for a fraction of the cost of major cloud providers.
• Vultr – A full-stack, enterprise-grade GPU platform for deploying and scaling AI applications worldwide.
• Masterworks – An exclusive platform that allows you to invest in shares of multi-million dollar paintings by artists like Banksy and Basquiat.
• Lemon.io – A marketplace for hiring top-tier, vetted software engineers from a global talent pool.
• LM Studio – An application to discover, download, and run local LLMs on your computer.