Good morning. It’s Wednesday, July 3rd.

In today’s email:

• Runway AI Video Now Available

• Apple’s 4M AI Model

• Apple’s OpenAI Board Seat

• 5 New AI Tools

• Latest AI Research Papers

You read. We listen. Let us know what you think by replying to this email.

Runway's Gen-3 Alpha AI Video Model Now Available For Monthly Fee

RunwayML’s latest update, Gen-3 Alpha, introduces enhancements in hyper-realistic video generation from text, image, or video inputs. Unlike earlier versions, Gen-3 Alpha now requires a paid subscription (starting at $12/month per editor, billed annually). It boasts improved capabilities, including precise key-framing, expressive human characters, and imaginative transitions, showcasing a leap in speed, fidelity, and motion quality.

Developed through a multidisciplinary team, Gen-3 Alpha initially focuses on text-to-video generation, with plans to expand to image-to-video and video-to-video functionalities in the future. Read more. 

Apple Launched a Public Demo of its '4M' AI Model. Here's Why it's a Big Deal.

Apple's notoriously guarded approach to research took a surprising turn with the public launch of their 4M AI model on Hugging Face Spaces.

The demo allows users to interact with 4M, demonstrating its abilities in generating images, performing object detection, and manipulating 3D scenes using natural language.

This public launch is a developer charm offensive. Apple's hoping to spark a vibrant ecosystem around 4M, attracting talent and fostering innovation. The timing is no coincidence either. With competitors making headlines in AI, Apple's move underscores their own quiet progress. In fact, Apple's stock has surged 24% since May 1st, partly fueled by their partnership with OpenAI.

But 4M's potential extends far beyond developer playgrounds. Its unified architecture suggests it could be woven into the very fabric of Apple's products, potentially supercharging Siri's intelligence or adding a whole new dimension to Final Cut Pro. However, concerns about data ethics arise due to AI's data-intensive nature. Read more.

Apple Poised to Get OpenAI Board Observer Role as Part of AI Pact

In a surprising move that throws shade on past rivalries, Apple has snagged an observer seat on OpenAI's board through a recent agreement, appointing Phil Schiller, head of Apple’s App Store and former marketing chief, to represent the tech giant. This partnership strengthens ties between Apple and OpenAI, marking a notable collaboration despite historical competition.

Schiller, as a board observer, will attend meetings without voting rights but gain insight into decision-making processes within OpenAI, similar to Microsoft Corp., OpenAI’s primary backer. The partnership aims to integrate OpenAI’s technologies into Apple products, including ChatGPT across iPhone, iPad, and Mac devices.

While Apple Intelligence enhances functionalities like article summarization and voice memo transcription, Apple’s alignment with OpenAI underscores the company’s pursuit of innovative AI solutions without a direct financial exchange, instead leveraging access to millions of potential users. Read more.

Etcetera: Stories you may have missed

Meta Is Changing Artificial Intelligence Labels After Real Photos Were Marked As AI

xAI's Grok 2: Elon Musk Unveils JARVIS-Inspired AI Assistant Arriving in August

Digit robots to be commercially deployed in GXO warehouses under industry-first RaaS model

Nvidia set to face antitrust charges in France over AI dominance

China to develop more than 50 new standards for AI sector by 2026

The Pixel 9’s ‘Google AI’ is like Microsoft Recall but a little less creepy

California advances unique safety regulations for AI companies despite tech firm opposition 

Anthropic looks to fund a new, more comprehensive generation of AI benchmarks

Robinhood snaps up Pluto to add AI tools to its investing app

New index using AI tools to measure U.S. economic growth in broader way

Google Cloud Nabs AWS, Microsoft Copilot Execs To Lead AI

Amazon proposes a new AI benchmark to measure RAG

China Plans Broader Push to Develop Tech Like Musk’s Neuralink

AI-based prison concept would complete sentences in just minutes

Meta drops '3D Gen' bomb: AI-powered 3D asset creation at lightning speed

Meta seeks to use generative AI in its metaverse, job listing shows

Deepfakes will cost $40 billion by 2027 as adversarial AI gains momentum

Open-source AI platform Sentient raises $85M co-led by Peter Theil’s Founder’s Fund

U.S. Awards $504 Million for ‘Tech Hubs’ in Overlooked Regions

Figma to temporarily disable 'Make Design' AI feature amid plagiarism concerns  

Motiff is an AI-powered UI design tool that enhances efficiency and collaboration with features like AI Toolbox and cloud integration.

ScienHub is an AI-enhanced LaTeX collaboration platform with language refinement, Git integration, and tools for high-quality academic papers.

Rapport combines AI, animation, and various tools to create, animate, and deploy Virtual Interactive Personalities (VIPs) on any platform.

Cre[ai]tion lite employs generative AI for visual design workflows, providing endless inspiration, dedicated design boards, and variation capture for designers.

Global Sites enables building static websites with AI-powered translations, supporting language switching, sitemap generation, and HTML validation in the boilerplate.

📄 MMEvalPro: Calibrating Multimodal Benchmarks Towards Trustworthy and Efficient Evaluation

📄 We-Math: Does Your Large Multimodal Model Achieve Human-like Mathematical Reasoning? ​​

📄 LLaRA: Supercharging Robot Learning Data for Vision-Language Policy

📄 DiffIR2VR-Zero: Zero-Shot Video Restoration with Diffusion-based Image Restoration Models

📄 MIRAI: Evaluating LLM Agents for Event Forecasting

📄 OmniJARVIS: Unified Vision-Language-Action Tokenization Enables Open-World Instruction Following Agents

📄 Wavelets Are All You Need for Autoregressive Image Generation

📄 RealTalk: Real-time and Realistic Audio-driven Face Generation with 3D Facial Prior-guided Identity Alignment Network

📄 Consistency Flow Matching: Defining Straight Flows with Velocity Consistency

Thank you for reading today’s edition.

Your feedback is valuable. Respond to this email and tell us how you think we could add more value to this newsletter.

Interested in reaching smart readers like you? To become an AI Breakfast sponsor, apply here.