🎓 University of America — Course Portal
🤖 Artificial Intelligence Week 1 of 14 ⏱ ~48 min BSc Year 1

Principles of Artificial Intelligence

The history, philosophy, and fundamental paradigms of AI — from Aristotle's logic to GPT-4. We explore what intelligence means, how humans attempted to replicate it in machines, and the major breakthroughs that shaped the field we study today.

AI101 — Week 1: Principles of Artificial Intelligence
AI-Generated Lecture · University of America
0:00 / 48:00
🎬
AI-Generated Video Ready Script: ~3,400 words (~48 min). Generate avatar video with Synthesia or HeyGen. Voice: ElevenLabs Rachel or Adam voice.
🎯 Learning Objectives
  1. Define artificial intelligence and explain why it is difficult to define precisely.
  2. Trace the historical development of AI from ancient philosophy to modern deep learning.
  3. Identify and explain the four major paradigms of AI (symbolic, connectionist, statistical, and hybrid).
  4. Describe the Turing Test and explain its philosophical implications and limitations.
  5. Implement a simple search-based AI agent in Python.
🤔 1. What Is Intelligence? What Is Artificial Intelligence?

Before we can study artificial intelligence, we must ask: what is intelligence? Philosophers, cognitive scientists, and AI researchers have wrestled with this question for centuries — and there is still no consensus.

Four Candidate Definitions of AI (Russell & Norvig) Systems that think like humans  |  Systems that think rationally  |  Systems that act like humans  |  Systems that act rationally. The field has pursued all four, and each leads to a different AI research program.

For this course, we'll work with this working definition: Artificial intelligence is the design of computational systems that can perceive their environment, reason about it, learn from experience, and take actions that maximize their chance of achieving goals.

Notice that this definition says nothing about consciousness, emotion, or even human-likeness. This is intentional. Much of modern AI is purely goal-directed and quite unlike human intelligence in its mechanisms.

📅 2. A Brief History of AI

AI is both ancient and new. The dream of thinking machines goes back thousands of years, but the scientific field is less than 80 years old.

Ancient Greece
Formal Logic (Aristotle, ~350 BCE)
Aristotle's syllogisms: the first formal system for reasoning from premises to conclusions. The foundation of symbolic AI.
1936
Turing Machine
Alan Turing proves any computable function can be computed by a simple tape-based machine. Establishes the mathematical foundation for computers and computation.
1950
Turing's "Computing Machinery and Intelligence"
Proposes the "Imitation Game" (later the Turing Test). First major philosophical argument for machine intelligence.
1956
The Birth of AI — Dartmouth Conference
McCarthy, Minsky, Shannon and others coin the term "Artificial Intelligence" and launch the field. Optimism is sky-high: human-level AI in 20 years, they said.
1969–1980
First AI Winter
Progress stalls. Early AI systems work only on toy problems. Funding dries up. The "combinatorial explosion" problem becomes clear.
1980s
Expert Systems
Rule-based systems encode human expert knowledge. MYCIN diagnoses blood infections. Major commercial interest — then another winter.
1986
Backpropagation (Rumelhart & Hinton)
Efficient algorithm for training multi-layer neural networks. Ignored for decades, then becomes the foundation of deep learning.
1997
Deep Blue Beats Kasparov
IBM's chess computer defeats the world champion. Symbolic AI's greatest triumph — but it uses no learning, just search and evaluation functions.
2012
AlexNet and the Deep Learning Revolution
Krizhevsky's deep CNN halves the ImageNet error rate. GPUs + big data + deep networks = modern AI. The field transforms overnight.
2017
Attention Is All You Need — Transformers
Google researchers publish the Transformer architecture. All of modern NLP, LLMs, and multimodal AI descends from this paper.
2022–Present
The Age of Foundation Models
GPT-4, Claude, Gemini, Llama — large language models achieve remarkable general capability. AI enters every domain of human activity.
🧩 3. Four Paradigms of AI

AI research has not followed a single path. Four major paradigms have each had their moment — and modern AI blends all four.

🔡

Symbolic AI (GOFAI)

Logic and rules. If-then reasoning. Knowledge bases. Expert systems. Good for structured reasoning, terrible at perception.

🧠

Connectionism

Neural networks inspired by the brain. Learn patterns from data. Dominant approach today — deep learning is here.

📊

Statistical AI

Probabilistic models, Bayesian networks, HMMs. Uncertainty-aware reasoning. Foundation of speech recognition and NLP.

🔀

Hybrid Approaches

Combine neural learning with symbolic reasoning. AlphaGo mixes RL, MCTS, and deep networks. The frontier today.

🎮 4. The Turing Test & Its Limits

In 1950, Alan Turing proposed a test for machine intelligence: could a human evaluator, conversing via text, distinguish a machine from a human? If not, the machine is intelligent by this criterion.

🎭 Turing Test — Example Dialogue
Human: Do you feel pain?
AI: That's a fascinating question. I process information about pain in humans from my training, and I can describe what pain does to people — but whether I "feel" anything is genuinely unclear to me. I represent states, but whether any of those representations involve something it's like to be me — that I honestly don't know.
Human: That sounds like a dodge.
Fair. Let me be more direct: I don't believe I experience pain. But I also can't rule it out. The honest answer to questions about machine experience is "I don't know, and neither does anyone else."
📌 Critical Thinking: The Turing Test measures behavioral indistinguishability, not intelligence per se. Philosopher John Searle's "Chinese Room" argument suggests that passing the Turing Test doesn't require understanding. Modern LLMs pass the Turing Test easily — but are they intelligent? This question is still unresolved.
💻 5. Your First AI Agent in Python

Let's implement a simple search-based AI agent. We'll build an A* solver for the 8-puzzle — one of AI's classic benchmark problems.

# AI101 — Week 1: A* Search Agent for 8-Puzzle import heapq def manhattan_distance(state, goal): """Heuristic: sum of Manhattan distances for each tile.""" dist = 0 for i, tile in enumerate(state): if tile != 0: # 0 is the blank goal_idx = goal.index(tile) dist += abs(i//3 - goal_idx//3) + abs(i%3 - goal_idx%3) return dist def get_neighbors(state): """Return all states reachable by one move.""" neighbors = [] blank = state.index(0) row, col = blank // 3, blank % 3 moves = [(-1,0),(1,0),(0,-1),(0,1)] # up, down, left, right for dr, dc in moves: new_row, new_col = row + dr, col + dc if 0 <= new_row < 3 and 0 <= new_col < 3: new_state = list(state) swap_idx = new_row * 3 + new_col new_state[blank], new_state[swap_idx] = new_state[swap_idx], new_state[blank] neighbors.append(tuple(new_state)) return neighbors def a_star(start, goal): """A* search: finds the shortest path from start to goal.""" heap = [(0 + manhattan_distance(start, goal), 0, start, [])] visited = set() while heap: f, g, state, path = heapq.heappop(heap) if state in visited: continue visited.add(state) path = path + [state] if state == goal: print(f"Solved in {g} moves! Explored {len(visited)} states.") return path for neighbor in get_neighbors(state): if neighbor not in visited: h = manhattan_distance(neighbor, goal) heapq.heappush(heap, (g + 1 + h, g + 1, neighbor, path)) return None # Example: solve from scrambled state to ordered start = (1,2,5, 3,4,0, 6,7,8) goal = (1,2,3, 4,5,6, 7,8,0) solution = a_star(start, goal)
📁 PROJECT 1Weight: 50% of AI101 Grade

Build a Game-Playing AI Agent

Implement an AI agent that plays a two-player board game (Tic-Tac-Toe, Connect 4, or Othello). Your agent must use search with a heuristic evaluation function. It should be unbeatable at Tic-Tac-Toe.

40%
Agent correctness & strength
30%
Code quality & documentation
30%
Analysis & reflection quality

📅 Due: 3 weeks · Submit GitHub + demo video via portal · Late: -5%/day

📝 Midterm & Final Preview
Midterm Q1 · Short Answer SA

Explain the distinction between "thinking like humans" and "acting rationally" as goals for AI systems. Which goal does modern deep learning pursue? Justify your answer.

Midterm Q3 · Algorithm Analysis Analysis

Given the 8-puzzle A* implementation from Lecture 1, trace the first 5 states expanded when solving the example puzzle. Show the f, g, and h values for each. Is the heuristic admissible? Prove it.

Final — Integrative Essay Essay

Write a 600-word essay comparing symbolic AI and connectionist AI. Use at least three historical examples. Argue for which paradigm you believe is more promising for achieving general intelligence, and why. Your argument must engage with at least one counterargument.

Final — Coding Problem Coding

Implement a Bayesian Naive Classifier from scratch in Python (no scikit-learn). Train it on the provided spam dataset and report precision, recall, and F1 score. Then: explain one scenario where Naive Bayes fails and why.