The hardest part is . Given a messy real-world scenario (e.g., "Help a lost tourist find a taxi while conserving battery"), how do you translate that into PDDL actions and MDP rewards? There is no single right answer. If you choose the wrong abstraction level (e.g., modeling every millimeter of movement), your planner will run until the sun dies.
And that is the mark of a great AI course: It doesn't just teach you algorithms. It teaches you to think like a planner. Have you taken CS5446 or a similar planning course? What was your "aha!" moment? Let me know in the comments. cs5446 ai planning and decision making
When most people think of "Artificial Intelligence," they picture a neural network recognizing a cat or ChatGPT writing a sonnet. But there is a quieter, more logical side of AI—the side that answers the question: "If the world looks like this , what sequence of actions should I take to make it look like that ?" The hardest part is