🌞 News

👉 Runway Aleph lets you change camera angles from your videos

Runway Aleph is a new in-context video model that enhances video editing and generation capabilities. It allows users to perform a variety of tasks, including adding, removing, and transforming objects in videos. Users can generate new camera angles, create seamless continuations of scenes, and apply different styles to their footage with simple prompts.

Additionally, Aleph enables users to change environments, alter the appearance of characters, and recolor elements in a scene, all while maintaining the integrity of the original footage. The tool also offers advanced features like motion application, relighting shots, and isolating subjects for green screen use. Early access will soon be available for Enterprise and Creative Partners, with broader access planned for all users.

👉 Samsung invests in Memories.ai to analyse 10M hours of videos

Memories.ai, a startup focused on improving video analysis, has secured $8 million in seed funding, led by Susa Ventures with participation from Samsung Next and other investors. The company aims to tackle the limitations of existing AI tools that struggle to analyze extensive video footage over long periods. By processing up to 10 million hours of video, Memories.ai provides a contextual layer with searchable indexing, tagging, and aggregation, catering primarily to marketing and security companies.

The startup's co-founders, Dr. Shawn Shen and Enmin Zhou, previously worked at Meta and are developing a solution that emphasizes on-device computing, which could enhance security applications by addressing privacy concerns. Memories.ai's technology aims to help users analyze trends in marketing videos and assess security footage for potential threats. The company plans to use the funding to expand its team and improve its video search capabilities, positioning itself against competitors in the evolving landscape of video intelligence.

👉 Google helps historians interpret ancient texts with AI

The Aeneas team from Google Deepmind has introduced a AI model designed to help historians contextualize ancient inscriptions, facilitating the interpretation and restoration of fragmentary texts. This model can process thousands of Latin inscriptions, rapidly identifying textual and contextual parallels that assist historians in their research. Aeneas is adaptable to various ancient languages and formats, expanding its utility across a range of historical evidence.

Developed in collaboration with several universities, Aeneas offers advanced capabilities such as searching for parallels, processing multimodal inputs, and restoring gaps in damaged texts. The model aims to integrate into existing research workflows, providing historians with interpretative suggestions and enhancing their ability to analyze historical documents. Aeneas is also open-sourcing its code and dataset to support further research and educational initiatives, reinforcing its commitment to advancing historical scholarship through technology.

👉 YouTube Shorts introduces new creation tools

YouTube Shorts has launched new creation tools aimed at enhancing user experience and creativity. One of the standout features, "Photo to Video," allows users to transform images from their camera roll into dynamic videos by adding movement or animation. This feature is rolling out for free in the US, Canada, Australia, and New Zealand, with plans to expand to additional regions later this year.

In addition, new generative effects will enable users to creatively manipulate their content, such as turning doodles into images or transforming selfies into unique video scenarios. These tools, powered by Veo 2, will be accessible through the Shorts camera and are set to reach more creators globally in the coming weeks.

👉 Figma Make now available to all users

Figma has officially launched Figma Make, its prompt-to-app coding tool, making it accessible to all users, not just those with a full subscription. This tool allows users to create working prototypes and applications using natural language descriptions, similar to other AI coding tools like Google’s Gemini Code Assist and Microsoft’s GitHub Copilot. While all users can experiment with Figma Make, publishing designs will still be limited to those on the Full Seat plan.

Additionally, Figma is implementing a new AI credit system that allocates a set number of credits for using the platform's AI tools, with limits varying by subscription tier. Users on lower-tier plans will have restricted access, while Full Seat users will enjoy more extensive capabilities. Figma Make also allows the inclusion of design references, enabling users to guide the AI in creating specific prototypes, enhancing the overall creative process.

 ⚡️ Flash Updates

· Google has introduced AI-generated store summaries in Chrome for U.S. shoppers, providing insights on store reputation regarding product quality, pricing, and customer service.

· Midjourney has announced new features , including the ability to create videos with specific start and end frames, looping videos, and video generation support on Discord. 

· Tencent has released the open-source Hunyuan3D World Model 1.0, enabling users to generate immersive and interactive 3D worlds from text or images, compatible with CG pipelines for editing and simulation.

· Pika labs released an early access AI-only social app that allows users to create and share transformed selfie videos with various sounds, styles, and effects.

· Microsoft has launched "Copilot Mode"  in its Edge browser, an AI-powered feature that assists users with web browsing by predicting their needs and performing tasks on their behalf.

· Elon Musk announced that his social media company X will revive the popular video-sharing platform Vine in "AI form," nearly nine years after its discontinuation.

👀 Creative picks

GEO Magazine and BETC use AI to revive extinct wildlife

GEO magazine, in partnership with BETC Fullsix, has launched an innovative project to visually resurrect five extinct animal species through AI-generated imagery. Using artificial intelligence, the campaign aims to create realistic representations of the couagga, Steller's sea cow, dodo, thylacine, and koala lemur, highlighting the urgent need for awareness around species extinction caused by human activity. Under the direction of science journalist Sylvie Redon-Clauzard, the project involved collaboration with renowned palaeontologists and biologists to ensure scientific accuracy.

The AI process was powered by Vermeer, a proprietary image generation tool developed by BETC in conjunction with Prose on Pixels. This technology facilitated a meticulous creative process, allowing the team to train the AI on a variety of historical data, including sketches and skeletons, to recreate the animals' appearances accurately. Tiphaine du Plessis, president of BETC Fullsix, emphasized that the campaign exemplifies a meaningful use of AI in journalism, while Myrtille Delamarche, editor-in-chief of GEO, noted that this project enables the magazine to present a compelling visual narrative on species that can no longer be photographed. The resulting images serve as a powerful reminder of the impact of human actions on wildlife and the importance of conservation efforts.

_{Credits: GEO / BETC Fullsix / Prose On Pixels}

🦄 Weekly visuals

🔥 Get your paper mahazine edition

THE MAHAZINE (AI magazine for creatives) [Numero 5, Spring 2025]

Be part of the future creative movement.

The mahazine is your exclusive invitation to join the inner circle. It's a highly curated, insider's guide to mastering AI for creative professionals, with research knowledge and exclusive insights you need to not just survive, but to redefine the future of your creativity.

What awaits you inside?

✍️ Become a prompt master: Learn the secrets to crafting prompts that unlock the full potential of AI, turning your visions into reality.

🖼️ Master different AI tools: Harness the power of MidJourney drafts to create content on the go and GPT-4o new edit features.

🎨 AI Architecture secrets: Discover the tools and techniques that will transform your architecture design workflow, giving you an unbeatable competitive edge.

🔥 Discover the hottest AI trends: Get exclusive access to the most innovative and impactful AI campaigns, so you're always one step ahead.

🦜 Unique visual explorations: Immerse yourself in our captivating visual experiments about birds, spring, medievial art and more.

Crafted by creatives, for creatives. The mahazine is more than just a magazine; it's a movement. Join us and become part of something extraordinary.

Limited printed number available. Don't miss out.

Best,

The Mahazine Team