GEN2050 SUMMER 2026
AGENTIC AI
What happens when you give AI a goal instead of just a question?
🤔 REGULAR AI vs AGENTIC AI
💬 Regular Chatbot AI
You ask a question
↓
AI gives one answer
↓
Done. That's it.
↓
AI gives one answer
↓
Done. That's it.
🤖 Agentic AI
You give it a goal
↓
AI makes a plan
↓
Uses tools on its own
↓
Checks if it worked
↓
Loops until the job is done!
↓
AI makes a plan
↓
Uses tools on its own
↓
Checks if it worked
↓
Loops until the job is done!
🔌 HOW AGENTS CONNECT TO TOOLS
An AI agent uses something called MCP (Model Context Protocol) to plug into real tools — like a universal adapter that lets it browse the web, check calendars, write code, send emails, and query databases all on its own:
🔄 THE AGENT LOOP
This is the cycle an AI agent repeats until the mission is complete:
GOAL
PLAN
ACT
OBSERVE
THINK
AGENT
LOOP
LOOP
🚀 PICK A MISSION
Choose a task and watch the AI agent work through it step by step:
AGENT'S TOOLBOX:
Select a mission above to begin
Agent log will appear here...
🧠 KEY IDEAS
Agentic AI is different because it doesn't just answer questions — it takes action to accomplish goals.
It can use tools like search engines, calculators, code editors, and APIs — just like you use apps on your phone.
When something goes wrong, it tries again with a different approach — just like you would.
The big idea: instead of being a tool you use, the AI becomes a co-worker that can handle multi-step tasks on its own. 🚀
It can use tools like search engines, calculators, code editors, and APIs — just like you use apps on your phone.
When something goes wrong, it tries again with a different approach — just like you would.
The big idea: instead of being a tool you use, the AI becomes a co-worker that can handle multi-step tasks on its own. 🚀