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:

• Google launches Gemini – Google has launched Gemini, its most capable and flexible AI model yet, which comes in three sizes (Ultra, Pro, and Nano) and is designed to be natively multimodal.
• Mistral AI releases Mixtral 8x7B – French AI startup Mistral AI has released Mixtral 8x7B, a high-quality, open-source sparse mixture-of-experts model (SMoE) that outperforms Llama 2 70B on most benchmarks and offers faster inference speeds.
• Pika Labs raises $55M and launches Pika 1.0 – AI video generator Pika has raised $55 million in funding and launched Pika 1.0, a new model that can generate and edit videos in various styles from text and image prompts.
• OpenAI announces GPT-4 Turbo and new developer products – At its first developer conference, OpenAI launched GPT-4 Turbo, which is more capable, cheaper, and has a 128K context window, alongside an Assistants API, and new multimodal capabilities.
• Elon Musk’s xAI launches Grok – Elon Musk’s xAI has officially launched its first AI model, Grok, a chatbot designed with a rebellious streak and real-time access to information from the X platform.
• Humane officially launches the Ai Pin – Humane has revealed its Ai Pin, a $699 screenless, AI-powered wearable device that clips to your clothing, projects information onto your hand, and operates via voice commands.
• The New York Times sues OpenAI and Microsoft for copyright infringement – The New York Times is suing OpenAI and Microsoft for copyright infringement, alleging that millions of its articles were used without permission to train their AI models.
• Stability AI releases Stable Video Diffusion – Stability AI has released Stable Video Diffusion, an open-source generative video model capable of creating short video clips from text or image prompts.
• Sam Altman returns as OpenAI CEO – Sam Altman has been reinstated as CEO of OpenAI with a new initial board, concluding a tumultuous period that saw him briefly fired from the company he co-founded.
• Amazon announces ‘Q’, an AI chatbot for businesses – Amazon has launched Q, a new generative AI-powered assistant specifically for businesses that can answer questions, generate content, and take actions based on a company’s internal data.
• Microsoft announces custom AI chips – Microsoft has unveiled its first custom-designed chips for AI workloads and cloud computing, named the Azure Maia AI Accelerator and the Azure Cobalt CPU.
• Internal warnings about an AI breakthrough preceded the ousting of OpenAI’s CEO – A letter from staff researchers to the OpenAI board about a powerful AI discovery called Q* (Q-star), which some feared could threaten humanity, reportedly contributed to the events leading to Sam Altman’s removal.
• Runway introduces Motion Brush for Gen-2 – Runway has launched Motion Brush, a new tool for its Gen-2 video model that allows users to add controlled motion to specific areas of a static image.
• The Browser Company raises $50 million – The Browser Company, creators of the Arc browser, raised $50 million to expand its vision of building “an internet computer” with integrated AI features.
• Meta introduces Audiobox – Meta has unveiled Audiobox, a new generative AI model for audio that can create voices, sound effects, and ambient noises from text prompts.
• Google releases a new AI model for weather forecasting – Google’s DeepMind has launched GraphCast, an AI model that can predict weather conditions up to 10 days in advance more accurately and much faster than traditional methods.
• Perplexity launches online LLMs that provide real-time, up-to-date answers – Perplexity has introduced ‘Online LLMs’, which integrate real-time web search capabilities directly into language models to provide current and accurate information.
• OpenAI quietly removes ban on military use from its policy – OpenAI has updated its usage policy, removing language that explicitly banned the use of its technology for “military and warfare” purposes.
• Satya Nadella says he would be open to Sam Altman returning to Microsoft if he leaves OpenAI again – Microsoft CEO Satya Nadella stated he is open to Sam Altman and his colleagues returning to Microsoft should they ever decide to leave OpenAI in the future.
• ElevenLabs launches a new multilingual voice generation model – AI voice company ElevenLabs has released its Multilingual v2 model, which can clone voices and generate speech in nearly 30 languages.
• U.S. and U.K. link up on AI safety – The United States and the United Kingdom have announced a partnership to jointly develop tests for the most advanced AI models, collaborating through their respective AI Safety Institutes.
• Apple is developing an AI-powered health coach – Apple is working on a new AI-powered health coaching service, codenamed Quartz, designed to help users stay motivated to exercise, improve eating habits, and sleep better.
• Introducing Devika – Devika is an open-source alternative to Devin that can understand high-level human instructions, break them down into steps, research relevant information, and write code to achieve the given objective.
• Ideogram introduces a ‘magic prompt’ feature – AI image generator Ideogram has added a “Magic Prompt” feature that automatically enhances and expands short user prompts to generate more creative and detailed images.
• Andrej Karpathy leaves OpenAI – Andrej Karpathy, a prominent AI researcher and founding member of OpenAI, has announced his departure from the company to pursue personal projects.
• AI could pass human bar for intelligence in 2024 – Experts predict that AI is on the verge of achieving Artificial General Intelligence (AGI) as large language models are expected to surpass the ‘human bar’ for intelligence this year.
• Getty Images partners with NVIDIA to launch Generative AI – Getty Images has collaborated with NVIDIA to create “Generative AI by Getty Images,” a new tool trained on Getty’s licensed library to produce commercially safe, copyright-indemnified images.
• Suno AI V3 – Suno has released V3 of its AI model, which can generate two-minute, radio-quality songs from simple text prompts.
• Reka Core: A new frontier multimodal model – Reka has launched Reka Core, its new flagship multimodal model that rivals leading models like GPT-4 and Claude 3 Opus in capabilities.
• Google plans to charge for new AI-powered search features – Google is considering charging for new premium features within its traditional search engine that are powered by generative AI, which would be a first for the company.
• Claude 3 running on a phone – A new project demonstrates that it’s possible to run Anthropic’s Claude 3 Haiku model entirely on-device on a phone.
• EU lawmakers approve the AI Act – The European Parliament has given final approval to the AI Act, a landmark law that sets comprehensive rules for artificial intelligence across the European Union.
• The AI Pin is here – MKBHD provides a detailed review of the Humane Ai Pin, concluding that the product is “the least impressive piece of tech I’ve ever tested.”
• Google adds image generation to Bard – Google’s Bard chatbot can now generate images for free, powered by the Imagen 2 model, in most countries around the world.
• Midjourney introduces character consistency – Midjourney has launched a new “Character Reference” feature that allows users to maintain consistent characters across multiple generated images.
• Leaked deck from an Andreessen Horowitz partner argues open-source AI will win – A leaked document from a Google researcher argues that open-source AI models are rapidly catching up to proprietary ones and will ultimately win due to their flexibility and community-driven innovation.
• Apple researchers create an AI model that can ‘see’ screen context – Apple has developed a new AI model named ReALM that can understand on-screen content and conversational context, potentially enabling a much smarter version of Siri.
• Intel and partners to build AI supercomputer with 1 million Intel CPUs – Intel is collaborating with Dell and other partners to construct an AI supercomputer powered by one million Intel Xeon CPUs.
• Google DeepMind’s ‘SimbA’ agent can beat top human players in poker – Researchers have developed an AI agent named SimbA (Single-model-based Bayesian-policy-search Agent) that can defeat top human players in no-limit Texas hold’em poker.
• What I learned from shipping an AI product in 2023 – Ethan Mollick shares key insights from developing and launching an AI-powered product, emphasizing the importance of user interaction and the unpredictable nature of AI applications.
• Introducing the TripoSR – A new open-source model has been released that can create high-quality 3D models from a single image in under a second.
• ChatGPT can now remember things you discuss to make future chats more helpful – OpenAI is testing a “memory” feature for ChatGPT that allows it to learn from past conversations to provide more relevant and personalized responses.
• Waymo is opening up its driverless ride-hailing service to the general public in Los Angeles – Waymo has launched its fully autonomous ride-hailing service to the public in Los Angeles, covering a 63-square-mile area from Santa Monica to Downtown LA.
• Open-source AI is about to have a field day – This article argues that the complexity and closed nature of top proprietary AI models create a significant opportunity for simpler, more transparent, and customizable open-source alternatives to thrive.
• Large language models cannot self-correct – A new study reveals that despite perceptions, large language models are not genuinely capable of self-correcting their own reasoning errors without specific external feedback.
• Why everyone is bad at writing AI prompts – This article explores the common difficulties users face in crafting effective AI prompts and suggests that the problem lies in the design of AI interfaces rather than user error.
• OpenAI Used Kenyan Workers on Less Than $2 Per Hour to Make ChatGPT Less Toxic – An investigation revealed that OpenAI utilized outsourced Kenyan laborers paid less than $2 an hour to label and filter toxic content, highlighting the human cost behind refining AI systems.
• VLOGGER – Google researchers have developed VLOGGER, an AI method that can generate controllable videos of people speaking and moving from just a single still photograph.
• Character.ai introduces a group chat feature – Character.ai now allows users to create group chats with multiple AI characters and other humans, enabling dynamic, multi-character interactions.
• AI video startup Magic unveils a superhuman video editor – Magic has launched an AI-powered video editing tool that uses natural language commands to perform complex editing tasks automatically.
• Code Llama 70B – Meta has released Code Llama 70B, its largest and best-performing version of the open-source coding assistant to date.
• Google Maps introduces new AI features – Google Maps is rolling out new generative AI capabilities that allow users to discover new places with conversational search and get more detailed navigation information.
• How AI could change computing forever – This article from Sequoia Capital provides a comprehensive overview of the generative AI landscape and predicts its transformative impact on creativity, software development, and various industries.
• How to build a startup in the new AI era – OpenAI CEO Sam Altman delivers a lecture at Stanford, offering his advice and insights on building successful startups in the current age of artificial intelligence.
• Microsoft is putting a Copilot key on new Windows 11 PCs – Microsoft is adding a dedicated Copilot key to the keyboard layout of new Windows 11 PCs, providing instant access to its AI assistant.
• Poe introduces a Mac app and 1 million-token context windows – The AI chatbot platform Poe has launched a native Mac application and is now supporting context windows of up to one million tokens for models like Claude 2.1.
• NVIDIA AI chatbot, Chat with RTX, runs locally on your PC – NVIDIA has released Chat with RTX, a demo application that allows users to run a personalized AI chatbot locally on their RTX-powered Windows PC.
• OpenAI is raising funds at an $80B+ valuation – OpenAI has completed a deal that values the company at $80 billion or more, allowing employees to cash out their shares.
• GitHub Copilot Enterprise is now generally available – GitHub’s Copilot Enterprise tier, which provides AI coding assistance tailored to a company’s own private codebase, is now available for all businesses.

Trending AI Tools:

• Luma Labs Dream Machine – A new, publicly available AI model for generating high-quality video from text and images.
• Stable Diffusion 3 Medium – The most advanced open-source text-to-image model from Stability AI, notable for its quality and ability to understand complex prompts.
• Codestral – A new 22B parameter, open-weight generative AI model from Mistral AI, specifically designed for code generation tasks.
• Krea – A suite of creative AI tools that allows for real-time generation and enhancement of images and video.
• Ploom – A new tool from Luma Labs that creates lifelike 3D animations and fly-throughs from a single image.
• MindStudio – A no-code platform that enables users to build and deploy custom AI applications for their business.
• Sora – OpenAI’s highly advanced text-to-video model that can generate realistic and imaginative scenes from text instructions (not yet publicly available).
• Perplexity Pages – A new feature from Perplexity that transforms search queries or topics into comprehensive, customizable, and shareable articles.
• Pika – An AI video platform that allows users to generate and edit videos in various styles from text and images.
• Kling – A high-fidelity text-to-video generation model from Chinese tech company Kuaishou, capable of producing long-duration, high-resolution videos (not yet publicly available).
• Suno – An AI music creation tool that generates songs, complete with vocals and instruments, from simple text prompts.
• Galileo – An AI tool that generates high-quality, editable UI designs for apps and websites from text descriptions.
• Splash Pro – An AI music generator that allows you to create original music and add custom AI-generated vocals.
• Magic Patterns – An AI-powered tool that generates production-ready UI components and frontend code from text prompts.
• Fine-Tuner.ai – A no-code platform designed to help users fine-tune large language models with their own data.
• Rivet – An open-source, visual programming environment for building and debugging complex AI agents and applications.
• Trieve – An open-source search infrastructure that allows developers to build advanced semantic search and RAG applications.
• Dola – An AI-powered calendar assistant that helps you manage your daily schedule and tasks directly within your chat app.
• Speechify – A text-to-speech application that can read aloud documents, articles, emails, and other text with natural-sounding voices.
• Kickresume – An AI-powered resume and cover letter builder designed to help users create professional job application documents quickly.
• Openlayer – An evaluation and testing platform for LLM applications that helps developers identify and resolve model failures.
• Keywrds.ai – An AI-powered tool designed to streamline and improve the keyword research process for SEO professionals.

Sponsors:

• Toloka – Toloka is an end-to-end LLM platform to build and deploy production-ready AI applications.
• Brilliant – Brilliant makes it fun to learn concepts in computer science, math, and AI through interactive lessons.
• Masterworks – Masterworks allows you to invest in blue-chip art from artists like Banksy and Basquiat, which has historically outperformed the S&P 500.
• The Rundown AI – The Rundown is the world’s fastest-growing AI newsletter, with over 600,000 readers from major companies like Apple, Google, and Meta.
• Kolena – Kolena is an ML testing platform that helps teams build more reliable models with curated tests for common and edge-case behaviors.
• RunPod – RunPod is a GPU cloud platform for AI/ML that has now integrated the SD3 API, enabling developers to generate high-quality images at a fraction of the cost.