Curriculum

Topic 01 · AI Disciplines

Artificial Intelligence

AI is the science of making machines that act rationally. From symbolic reasoning in the 1950s to neural networks today, it is the broadest category in this curriculum.

The idea

Artificial Intelligence began as a formal academic discipline at the 1956 Dartmouth Conference. The original vision was purely symbolic: encode human knowledge as logical rules and let a computer derive conclusions. Expert Systems in the 1980s showed this could work in narrow domains—MYCIN correctly diagnosed blood infections—but the knowledge-acquisition bottleneck was brutal. You had to interview experts and hand-encode every rule.

The Turing Test (1950) proposed a behavioural criterion: if a machine's conversation is indistinguishable from a human's, it is intelligent. Critics rightly point out that mimicking language is not the same as understanding it, but the test shaped two decades of NLP research.

Modern AI blends three approaches: symbolic reasoning for structured domains, statistical learning for pattern recognition, and neural networks for end-to-end representations. The biggest practical systems—AlphaFold, GPT-4, AlphaCode—are all hybrids at some level.

P(A|B) = P(B|A)·P(A) / P(B)

Bayes' theorem—the foundation of probabilistic reasoning in AI.

What to hold onto

  • Symbolic vs. Sub-symbolicRule-based systems are interpretable but brittle. Neural networks are flexible but opaque. The field has swung between them multiple times.
  • Search as AIA* pathfinding, minimax, Monte-Carlo tree search—AI often reduces to searching a state space for a goal state.
  • The No-Free-Lunch TheoremNo algorithm outperforms random search on all problems. Specialisation is mandatory; there is no universal AI solver.

Live: A* Pathfinder Simulator

Click cells to toggle walls. Drag or watch A* explore states with Manhattan heuristic $f(n) = g(n) + h(n)$.

● Start ● Goal ■ Path Open: 0 | Closed: 0

Quick answers

What is Artificial Intelligence?

Artificial Intelligence is the field of computer science focused on building systems that perform tasks requiring human-like reasoning, perception, or decision-making. It encompasses symbolic AI, machine learning, and neural networks, spanning applications from search algorithms to large language models.

What is the Turing Test?

The Turing Test, proposed by Alan Turing in 1950, evaluates machine intelligence by determining whether a human evaluator can distinguish a machine's responses from a human's in natural conversation. Passing the test suggests human-level language capability.