🤵🏻‍♂️Meta's Multimodal Masterpiece

May 22, 2024

Reading time: 5 minutes  

Can you help us, please?

Last week, we asked if you'd be interested in a free prompt guide, and the response was incredible! We're thrilled to share that we're already working on it. Now, we need your help to make it as useful as possible.

Login or Subscribe to participate in polls.

Today we will discuss-

  • 🔥Meta's multimodal AI innovation

  • ✨Adobe's generative remove feature

  • 🏫US educators get free AI assistant

  • ⚙️9 amazing AI tools you might not have heard of

All this and more - Let's dive in!

👩‍🍳What’s cooking in the newsroom?

Meta previews Chameleon, the next-generation AI for multimodal tasks

🤹🏻‍♀️News - Meta has provided a sneak peek into its upcoming family of AI models, dubbed Chameleon. This new line of models is engineered to manage various types of data simultaneously.

Although not officially released yet, Meta's initial experiments indicate that Chameleon excels in tasks like image captioning and visual question answering (VQA), while also demonstrating strong performance in text-only tasks.

🌟What's special about it? When creating models that understand different types of data like images and text, there are two main approaches. One way is to stitch together separate models for each data type, but this can make it hard for the models to work together smoothly. 

Chameleon, on the other hand, is like a special kind of multitasking model. It's built to handle a mix of data types, like pictures, words, and even code, all at once. It does this by transforming pictures into small chunks of data (tokens) and using a shared language that all the different types of data can understand. This makes it easier for Chameleon to understand and work with both pictures and words together, without getting confused.

🦎 In conclusion - According to the researchers behind it, Chameleon is a significant advancement towards achieving unified foundation models that can effectively reason and create multimodal content. It will be interesting to see it lives up to the hype though. 

Adobe Lightroom gets Firefly AI-powered 'Generative Remove' editing feature

👨🏻‍💻News - Adobe has added a new feature called "Generative Remove" in Lightroom, which is powered by its Firefly AI model. 

This new tool, available on Lightroom's mobile app, web client, and desktop application, is touted as the "most powerful removal yet" tool by Adobe.

Adobe claims that with generative remove, users can delete unwanted objects from any photo, regardless of the complexity of the background. The company asserts that this feature can handle even the most intricate backgrounds and surroundings, including tasks like removing stains from patterned shirts, smoothing wrinkles from tablecloths in food photography, and eliminating unwanted reflections in water.

Furthermore, according to the company, the feature intelligently matches the removed area with "pixel-perfect generations," resulting in high-quality, realistic, and impressive outcomes.

🪄See also - Adobe has also announced that the AI-powered Lens Blur tool in Lightroom is now available to all users, along with new presets. This feature enables users to instantly apply blur effects to modify the appearance of any photo. 

Microsoft and Khan Academy offer free AI assistant to educators across the US

👩🏻‍🏫News - Microsoft has teamed up with Khan Academy to offer a free generative AI assistant, Khanmigo for Teachers, to all US educators. 

This tool will help teachers prepare lessons, analyze student performance, plan assignments, and pursue their own professional development.

Microsoft and Khan Academy are also planning to provide AI-powered math tutoring for students using Microsoft's new open-source small language model, Phi-3. Since small language models are more targeted and focus on specific data sets, the AI math tutor will offer grade-specific guidance and personalized sessions for students and teachers. The initial focus will be on middle school, with plans to expand to other grades.

👩🏻‍🎓What's more? The collaboration between the tech giant and the education nonprofit will serve to enhance Microsoft's Copilot AI by integrating Khan Academy content into it. Additionally, Khan Academy content will be incorporated into Microsoft Teams for Education, a free communication app for students and teachers.

🙆🏻‍♀️What else is happening?

😺Did you know?

  • China leads the market of AI deployment, with 58% of companies in China already deploying AI.

  • In 2020, researchers at Cornell University developed a self-healing robot that could detect and repair damage to its own body. The robot was made of a soft, flexible material that was embedded with sensors.

  • AI is expected to contribute $15.7 trillion to the global economy by 2030. This is more than the current output of India and China combined.

👩🏼‍🚒Discover mind-blowing AI tools

  1. PicturePerfectAI - An AI-powered tool that allows users to transform their photos into unique avatars ($11.99 one-time payment)

  2. Notably - An AI-powered research platform that helps users analyze and understand research data more efficiently ($25/month)

  3. Dewey - An AI accountability buddy that helps users stay organized and productive (Free up to 3 open tasks at a time) 

  4. Entry Point - A platform that allows businesses to fine-tune AI models without any coding ($49/month) 

  5. YaatriAI - A travel planning platform that provides personalized itineraries based on users' preferences and budgets (Free)

  6. Ticket AI - A Discord bot that automates customer support by using AI to answer support tickets ($19/month)

  7. Ask Jules - A book discovery companion that helps users find their next book and answers book-related questions (Free)

  8. Krizmi - An interactive learning platform that offers auto-generated flashcards and quizzes to help students retain and test their knowledge (Free)

  9. Capsule - An AI-powered video editor designed for enterprise content and marketing teams ($99/month)

📓AI comic books

  • Stargazer Episode Zero (2022) - A short sci-fi comic book that was illustrated using Midjourney, an online image generator. The story retells the experiences of American soldiers in Vietnam when an extraordinary cosmic anomaly emerges on their path.

  • The Bestiary Chronicles (2022) - A four-part comics series illustrated entirely using Midjourney, with a few minor adjustments made via Photoshop. The series has been praised for its beautiful artwork, its imaginative take on the fantasy genre, and its innovative use of AI.

  • Goats (2022) - A comic book by filmmaker Elvis Deane, created using Midjourney. The story follows Julia, a young financial analyst who finds herself drawn into the mysterious world of the supernatural after moving to London.