Hello Niuralogists!

‍

Welcome to this week’s edition, where we dive into the latest advancements in artificial intelligence. We’ll explore how these innovations are shaping various aspects of our lives—from the workplace and business to policies and personal experiences. This issue highlights some fascinating updates, including issues around the use of copyright data to train AI, debates on the role of AI in art, new features of tech giants’ AI models, and more.

‍

For more in-depth coverage, keep reading… 

‍

Former OpenAI Developer Speaks Out Against AI’s Use of Copyrighted Data

Suchir Balaji, a former OpenAI employee, has spoken out against practices involving the use of copyrighted data for training AI models like GPT-4. He left the company in 2023 due to ethical concerns surrounding the use of such data, arguing that OpenAI’s reliance on such data without authorization violates copyright laws and undermines the sustainability of original content creators. Despite OpenAI’s claim that its practices fall under "fair use," Balaji believes the AI-generated content too closely mirrors the original works. His stance has sparked legal battles, calling for increased regulation to protect intellectual property in the AI age. This is the first time an ex-employee of an AI company has spoken out against current practices.

‍

‍

Google Open-Sources SynthID to Boost Transparency in AI-Generated Content

Google has made DeepMind’s SynthID open-source, allowing other companies and developers to use it. SynthID is a tool designed to watermark and identify AI-generated content, including images, audio, text, and video. The watermarks are embedded directly into the content, invisible to humans but detectable by AI systems. SynthID aims to address concerns about misinformation and ensure trust in AI-generated material. While it’s not a complete solution, it provides a crucial step in responsible AI use, and Google hopes it will become the industry standard for AI generated content. Currently in beta, SynthID is being integrated into various Google products and available to developers through platforms like Vertex AI.

‍

Runway Unveils New Tool for Expressive and Easy AI Animations

Runway’s Act-One simplifies the creation of AI-generated video by using a single video input to animate characters, unlike traditional methods that require complex motion capture equipment or manual rigging. It captures eye-lines, micro-expressions, and realistic emotions, offering creators more control over animated performances. This tool allows users to produce cinematic-quality animations with just a basic camera, enabling a broader range of artists to create high-quality content. Act-One also prioritizes safety with content moderation and public figure detection features, making it accessible and secure for creators.

‍

‍

Apple’s Developer Beta Unveils New AI Capabilities for Siri, Image Generation, and More

Apple's latest developer beta, including iOS 18.2 and macOS Sequoia 15.2, introduces new Apple Intelligence features. These include integrated tools for writing, image cleanup, and summaries, alongside ChatGPT integration for Siri and compositional tasks. Users can generate custom emojis, transform sketches with Image Wand, and create AI-driven images using prompts in Image Playground. Visual Intelligence, similar to Google Lens, enhances photo analysis and text recognition. Apple emphasizes safety, with licensed content for image generation and safeguards for ethical concerns in AI-generated media. Public rollout starts next week.

‍

Over 11,500 Creatives Demand Ban on AI Training with Unlicensed Art

Over 11,500 creative professionals have signed an open letter demanding a ban on the use of human art to train AI without permission. The letter highlights the unlicensed exploitation of artists' work as a threat to livelihoods, sparking ongoing legal and ethical concerns. While some publishers have struck deals with AI companies, others, like The New York Times, have filed lawsuits. The letter emphasizes the urgent need for regulation to protect creators as AI tools increasingly generate media using their likenesses without consent.