• The AI Pulse
  • Posts
  • 🤖 Microsoft’s Rebuilding the Internet for Agentic AI

🤖 Microsoft’s Rebuilding the Internet for Agentic AI

PLUS: NVIDIA Helps Humanoid Robots See, Act, and Do Just Like Humans

Author

James Roesser
May 20, 2025

Welcome back AI enthusiasts!

In today’s Daily Report:

  • 👷Microsoft’s Rebuilding the Internet for Agentic AI

  • 🦾NVIDIA Helps Humanoid Robots See, Act, and Do

  • 🛠Trending Tools

  • 🥪Brief Bites

  • 💰Funding Frontlines

  • 💼Who’s Hiring?

Read Time: 3 minutes

🗞RECENT NEWS

MICROSOFT

👷Microsoft’s Rebuilding the Internet for Agentic AI

At Microsoft Build 2025, Microsoft CEO Satya Nadella announced his comprehensive strategy to build what he calls the “Open Agentic Web.”

Key Details:

  • Microsoft provided more than 50 announcements spanning across the company’s entire product portfolio, all focusing on advancing the adoption of AI Agents.

  • The “Open Agentic Web” is a bold vision where AI Agents will require a complete rethinking of how websites, sensitive datasets, and authentication mechanisms are structured and accessed on the Internet.

  • AI Agents are starting to change how we interact with the Internet by acting on our behalf. They can autonomously operate across websites to browse, book, and buy without us lifting a finger.

  • While humans navigate websites visually, AI Agents navigate them structurally. In other words, we read pixels while they read code. The digital world has been optimized for humans. Now, it’ll have to be restructured to support AI Agents.

Why It’s Important:

  • Today’s websites prioritize visuals over underlying code, making it harder for AI Agents to function effectively. Until websites adapt to AI Agents, today’s underlying code formats of webpages will limit Agentic AI adoption.

  • Websites designed for AI Agents will prioritize MAchine-Readable Formats (MARF) over visually optimized User Interfaces (UIs).

🩺 PULSE CHECK

Are AI Agents a gateway to creating a Utopia or Dystopia?

Vote Below to View Live Results

Login or Subscribe to participate in polls.

NVIDIA

🦾NVIDIA Helps Humanoid Robots See, Act, and Do

NVIDIA just launched “Isaac GR00T N1.5,” a general-purpose foundation model for humanoid robots to seamlessly navigate, adapt, and interact with the real world.

Key Details:

  • It’s a Vision-Language-Action Model (VLA) that helps humanoid robots understand the real world through text and images to execute actions.

  • VLAs consist of four components:

    1. Text Encoder: Converts text into numbers.

    2. Image Encoder: Converts images into numbers.

    3. Information Fuser: Creates unified numbers that contain text and images.

    4. Action Decoder: Converts the unified numbers into actions or motor commands that the humanoid robot can understand.

  • “Issac GR00T N1.5” enables humanoid robots to understand the world like we do with our eyes and perform actions like we do with our bodies.

Why It’s Important:

  • The biggest challenge in robotics is generalization: the ability of humanoid robots to execute actions with new objects in new environments.

  • It’s not enough for an industrial humanoid robot to pick up a specific “box,” “crate,” or “pallet” in a specific warehouse. It must understand the general concept of a “box,” “crate,” or “pallet” so it can operate in countless warehouses.

🛠TRENDING TOOLS

⚡️TaoPrompt creates professional prompts fast.

📸imgEdify generates, edits, and transforms images for free.

❄️Brizy lets AI build a professional website for you in seconds.

💨Blaze creates, repurposes, and optimizes marketing content.

💻Cursor is an AI-powered code editor that predicts your next edit.

🥪BRIEF BITES

xAI’s Grok 3 can now generate bar charts, pie charts, line graphs, and scatter plots to visualize complex datasets in seconds.

China recently launched the first 12 satellites of a 2,800-satellite AI-powered three-body computing constellation that’ll process data directly in space.

Research Lead at OpenAI Jerry Tworek explained that GPT-5 will unify all of OpenAI’s AI-powered features, including Canvas, Operator, Deep Research, and Advanced Voice Mode.

Researchers at the University of London (UoL) discovered that AI Agents can make group decisions without human intervention, even persuading one another to change their minds.

💰FUNDING FRONTLINES

  • Kovr.ai raises a $3.6M Seed Round to transform Cyber Compliance Automation.

  • Origin secures a $21M Series A to build the world’s first Global Benefits Intelligence Platform.

  • Miter closes a $38M Series B to modernize how contractors manage payroll, HR, and field operations.

💼WHO’S HIRING?

  • NVIDIA (Santa Clara, CA): Compute Systems Software Intern, Fall 2025

  • Capital Rx (Remote): Analyst, Data Exchange, Entry-Level

  • Airwallex (San Francisco, CA): Software Engineer, GenAI, Mid-Level

  • Gusto (San Francisco, CA): Principal AI Software Engineer, Senior-Level

📒FINAL NOTE

FEEDBACK

How would you rate today’s email?

It helps us improve the content for you!

Login or Subscribe to participate in polls.

❤️TAIP Review of The Day

“Good layout with up-to-date AI info.👏”

-Will (1️⃣ 👍Nailed it!)
REFER & EARN

🎉Your Friends Learn, You Earn!

You currently have 0 referrals, only 1 away from receiving 🎓3 Simple Steps to Turn ChatGPT Into an Instant Expert.