• Data Pragmatist
  • Posts
  • OpenAI adds watermarks to AI-generated images using DALL-E 3

OpenAI adds watermarks to AI-generated images using DALL-E 3

Tech and AI Developments in a Glance

Author

Ishaan Mahajan
February 09, 2024

🌐Tech and AI Developments in a Glance

🤖 Apple releases a new AI model

🖼️ OpenAI adds watermarks to AI-generated images using DALL-E 3

📜 AI reveals hidden text of 2,000-year-old scroll

🖼️ Meta to start labeling AI-generated images

🤖 Apple releases a new AI model LINK

  • Apple has released “MGIE,” an open-source AI model for instruction-based image editing, utilizing multimodal large language models to interpret instructions and manipulate images.

  • MGIE offers features like Photoshop-style modification, global photo optimization, and local editing, and can be used through a web demo or integrated into applications.

  • The model is available as an open-source project on GitHub and Hugging Face Spaces.

🖼️ OpenAI adds watermarks to AI-generated images using DALL-E 3 LINK

  • OpenAI is implementing watermarks, including an invisible metadata component and a visible CR symbol, in DALL-E 3 images to support the standards set by the Coalition for Content Provenance and Authenticity (C2PA).

  • The new watermarking feature, which slightly increases image sizes and has negligible latency effects, allows users to verify the AI tool used for creating images.

  • Despite the initiative, OpenAI acknowledges the limitations of watermarks in combating misinformation, noting that C2PA’s metadata can be easily removed or lost, especially on social media platforms that often strip metadata from uploads.

One Thing You Should Watch

How AI Could Empower Any Business | Andrew Ng | TED

Expensive to build and often needing highly skilled engineers to maintain, artificial intelligence systems generally only pay off for large tech companies with vast amounts of data. But what if your local pizza shop could use AI to predict which flavor would sell best each day of the week? Andrew Ng shares a vision for democratizing access to AI, empowering any business to make decisions that will increase their profit and productivity. Learn how we could build a richer society – all with just a few self-provided data points.

📜 AI reveals hidden text of 2,000-year-old scroll LINK

  • A group of classical scholars, assisted by three computer scientists, has partially decoded a Roman scroll buried in the Vesuvius eruption in A.D. 79 using artificial intelligence and X-ray technology.

  • The scroll, part of the Herculaneum Papyri, is believed to contain texts by Philodemus on topics like food and music, revealing insights into ancient Roman life.

  • The breakthrough, facilitated by a $700,000 prize from the Vesuvius Challenge, led to the reading of over 2,000 Greek letters from the scroll, with hopes to decode 85% of it by the end of the year.

🖼️ Meta to start labeling AI-generated images LINK

  • Meta is expanding the labeling of AI-generated imagery on its platforms, including content created with rivals' tools, to improve transparency and detection of synthetic content.

  • The company already labels images created by its own "Imagine with Meta" tool but plans to extend this to images generated by other companies' tools, focusing on elections around the world.

  • Meta is also exploring the use of generative AI in content moderation, while acknowledging challenges in detecting AI-generated videos and audio, and aims to require user disclosure for synthetic content.