💻 Claude Code goes Web

Good morning, AI enthusiast. Anthropic just brought its viral coding assistant to the web, taking direct aim at GitHub Copilot, Cursor, and OpenAI’s code tools.

In today’s AI newsletter:

• Anthropic Brings Claude Code to the Web

• Deepseek’s OCR Lets AI Handle Longer Documents

• Woman Wins Court Case Using ChatGPT

• How to generate free cinematic AI videos with realistic motion and facial detail

• AI tools & more

AI NEWS

💻 Anthropic Brings Claude Code to the Web

Anthropic launched a web app for its AI coding assistant Claude Code, letting developers create and manage coding agents directly from a browser.

• Claude Code is now available to Pro ($20/mo) and Max ($100–200/mo) users via Claude or the Claude iOS app

• Anthropic claims 90% of Claude Code itself is written by its AI models

• Developers can spin up autonomous coding agents that work independently across multiple platforms

• The tool has grown 10× in users since May, now generating $500M+ annually

This launch reflects a major shift in software development: engineers are increasingly managing AI coding agents rather than writing code line-by-line. It also positions Anthropic to compete directly with GitHub Copilot, OpenAI, and Google in the AI coding space.

AI MEMORY

📝 Deepseek’s OCR Lets AI Handle Longer Documents

Chinese AI company Deepseek has developed an OCR system that compresses image-based text, allowing language models to process much longer documents without hitting memory limits.

• Converts text into images, reducing compute needs while retaining 97% of information

• Combines Meta’s SAM, OpenAI’s CLIP, and a 16× token compressor for efficiency

• Supports ~100 languages and various document types, including diagrams and formulas

• Can process 33 million pages per day on modern hardware

Deepseek-OCR drastically improves efficiency in text recognition, enabling AI to work with larger datasets and longer contexts. Its ability to compress documents while preserving nearly all information can transform AI training, knowledge management, and document-heavy applications.

AI LAW

⚖️ Woman Wins Court Case Using ChatGPT

A CalifoA woman in California successfully used ChatGPT and Perplexity to overturn her eviction notice and avoid tens of thousands in penalties, representing herself in court.

• AI tools identified procedural errors, suggested actions, and drafted court responses for her.

• Other litigants, like Staci Dennett, have also leveraged AI to settle disputes and improve legal arguments.

• Experts warn AI can be misleading; fabricated case citations have led to fines and sanctions for some attorneys.

The case highlights the growing use of AI in the legal system, showing its potential to empower individuals, but also the serious risks of relying on unverified AI-generated advice.

HOW TO AI

🎨 How to generate free cinematic AI videos with realistic motion and facial detail

In this tutorial, you will learn how to generate cinematic, photorealistic AI videos using Waver 1.0, ByteDance’s new on-device video generation model that outperforms tools like Veo 3, Kling, and Midjourney, and does it completely free through Discord.

🧰 Who is This For

• AI creators experimenting with video generation

• Filmmakers and storytellers testing AI cinematography

• Educators or content creators seeking high-quality free visuals

• Anyone curious about next-gen video AI models

STEP 1: Join Discord Server

Go to the official AI Discord server. Click the “Join” button when prompted, and you’ll automatically enter their main workspace, which looks like a chat interface similar to WhatsApp or Slack.

Once you’re in, locate the sidebar on the left. You’ll see multiple channels under categories like “t2v-generate-horizontal” and “t2v-generate-vertical.” These are the channels used to generate videos.

STEP 2: Access the Video Generation Channel

Select any of the text-to-video channels such as “t2v-generate-horizontal” for landscape videos or “t2v-generate-vertical” for mobile content. Each channel allows users to type prompts that the AI will use to create a video.

Before writing your first prompt, ensure the message input box at the bottom shows the text placeholder, this means the bot is active and ready to generate.

STEP 3: Generate Your Video

Type your video description in this format:

/generate_video prompt: “Close-up of a 3D butterfly landing on a flower, spreading wings, smiling, colorful cartoon environment for kindergarten video.”

After typing your prompt, press Enter. The system will process your request, and within a minute, you’ll see your generated video appear directly in the chat.

STEP 4: Preview and Download Your Video

To preview the generated clip, click the Play icon in the center of the thumbnail. If you want to download it, select the Download button located in the lower-left corner of the video message.

Your cinematic AI video will be saved instantly in high-quality MP4 format on your system, no watermark, no hidden cost, completely local.

AWS says services have “returned to normal operations” after an outage in its US-EAST-1 region, first reported at 3:11am ET, that took down major websites.

CleanSpark closed up 4.48% after the bitcoin miner said it would expand into AI infrastructure, hiring Humain president Jeffrey Thomas as SVP of AI data centers.

Researchers found that LLMs can get “brain rot” from bad web data, losing reasoning, memory, and safety skills that don’t fully recover even after retraining.

Anthropic launched Claude for Life Sciences with new lab tools, scientific connectors, and better biomedical performance.

⚙️ Claude Code: Anthropic’s coding assistant, now on web

🎥 Veo 3.1: Google’s new and improved video AI model

🗂️ Skills: Claude’s new folder system for adding skills

💨 Claude 4.5 Haiku: Anthropic’s new fast and affordable AI model

THAT’S IT FOR TODAY

Thanks for making it to the end! I put my heart into every email I send, I hope you are enjoying it. Let me know your thoughts so I can make the next one even better!

See you tomorrow :)

- Dr. Alvaro Cintas