#6 Generative vs Agentic AI

ℹ️ Overview

Omar Shanti provides an insightful explanation of the difference between generative AI and agentic AI, illustrating how these advancements are reshaping the AI landscape. He uses relatable metaphors and real-world examples to highlight how agentic systems represent the next phase of AI evolution, where models are not only capable of generating content but also executing actions autonomously. This shift is accelerating productivity and enabling more proactive, assistant-like interactions.

📒 Key Takeaways

• Defining Generative AI

  • Generative AI models, like an earlier version of ChatGPT, rely on pre-trained data to generate responses.
  • Users must supply data for analysis, as these models don’t independently identify or request what they need.
    
    

• What is Agentic AI?

  • Agentic AI introduces the ability to plan and act autonomously.
  • It behaves like a librarian that identifies and retrieves the resources required to answer a query, instead of being handed a book directly.
  • These models can execute tasks such as opening browsers, navigating to specific websites, and interacting with software systems.
  • Agentic AI integrates two key components:
    1. Planning Arm – Determines what data or actions are required.
    2. Executive Arm – Executes tasks or code to fulfill the query.
       
       

• Agentic AI in Practice

  • Example: In recruitment (RPO world), an agentic AI could pull real-time market research, analyze candidate pools, track past communication, and even proactively engage with candidates.
  • Tools like Claude's Model Context Protocol (MCP) enable these systems to interact with file systems, browsers, and other services seamlessly.
    
    

• Impact on Productivity

  • Agentic AI accelerates workflows significantly:
    • Tasks that once took weeks can now be completed in hours.
    • Hackathon examples showed deliverables completed in 6–8 hours instead of 6 weeks.
  • Results depend on familiarity with tools and the complexity of the task.
    
    

• Collaboration with Software Programs

  • Agentic AI reduces human involvement by collaborating with other software systems, automating tasks like scheduling, data retrieval, and execution.
  • This creates a dynamic assistant-like experience, where AI can manage tasks autonomously and only notify users when necessary.