• FavTutor
  • Posts
  • Claude AI can now run and write JS code

Claude AI can now run and write JS code

Including the latest AI news of the week

October 28, 2024

Hello, AI Enthusiasts!

Welcome to FavTutor’s AI Recap! We’ve gathered all the latest and important AI developments from the past week in one place, just for you.

In Today’s Newsletter: 😀 

  • Claude AI can now run and write JS code

  • OpenAI's new AI model generates images 50x faster

  • Google DeepMind open-sources AI text watermarking for Gemini

  • Midjourney tests new AI image editor for uploaded images

  • Google's rumored Gemini 2.0 could support LLM stagnation thesis

Anthropic
👨‍💻 Claude AI can now run and write JS code

Anthropic’s new built-in feature for Claude.ai enables Claude to write and run JavaScript code. Claude can now process data, conduct analysis, and produce real-time insights.
Insights for you:

  • Anthropic equips Claude.ai with a JavaScript code sandbox that directly enables the application's complex data analyses and calculations, including visualizations.

  • Because the generated code runs directly in the chat tool, the feature can make AI coding more accessible, especially for non-coders.

  • The tool acts as an integrated code environment where Claude can perform mathematical calculations, analyze data sets, and test different approaches before delivering results.

Learn AI in 5 Minutes a Day

AI Tool Report is one of the fastest-growing and most respected newsletters in the world, with over 550,000 readers from companies like OpenAI, Nvidia, Meta, Microsoft, and more.

Our research team spends hundreds of hours a week summarizing the latest news, and finding you the best opportunities to save time and earn more using AI.

OpenAI
🖼️ OpenAI's new AI model generates images 50x faster

OpenAI introduces a new method that dramatically simplifies and accelerates the training of AI image models. The new sCM method makes training these models more stable and scalable.

Insights for you:

  • OpenAI has developed a new method called sCM that makes AI image generation models faster and simpler to train.

  • This enables high-quality image creation in just two computation steps instead of the many more steps previously required.

  • The largest sCM model, with 1.5 billion parameters, generates images in 0.11 seconds on an A100 GPU, which is 50 times faster than traditional diffusion models.

Google
💡 Google DeepMind open-sources AI text watermarking for Gemini

Google DeepMind has added text recognition to its SynthID AI watermarking technology. The company is integrating this feature into its Gemini models and releasing it as an open-source project.

Insights for you:

  • The company has integrated this feature into its Gemini models and released it as an open-source project.

  • SynthID for Text works by subtly adjusting token probability scores during text generation. This process creates a watermark pattern without affecting the output's quality or creativity, according to Google Deepmind.

  • Google DeepMind has made SynthID available through partnerships with Hugging Face and as part of its Responsible Generative AI Toolkit.

Midjourney
💡 Midjourney tests new AI image editor for uploaded images

Midjourney is testing two new features for its AI image generator: an image editor for uploaded images and an image restructuring feature that allows users to change materials, surfaces, and lighting.

Insights for you:

  • Midjourney launches two new features: an updated image editor for uploaded images, including an image transformation tool for changing materials, surfaces, and lighting.

  • The features integrate with all existing advanced features such as style references, character references, and personalized models, although some issues remain with small area edits and body proportions.

  • Initial access is limited to annual subscribers, monthly subscribers in the past 12 months, and users who have generated more than 10,000 images.

Google
🤔 Google's rumored Gemini 2.0 could support LLM stagnation thesis

The Verge's sources suggest that Gemini 2.0 may not deliver the significant performance improvements that the team, led by Deepmind founder Demis Hassabis, had anticipated.

Insights for you:

  • Google plans to release its new flagship LLM, Gemini 2.0, in December.

  • Major tech companies like Google, Meta, and xAI are investing heavily in new talent and computing resources to develop specialized reasoning models.

  • These models focus more processing power on real-time thinking rather than pre-training, aiming to find new ways to improve AI performance.