Foundations of AI: Cognitive Science Influences — Beginner-Friendly Notes
1. What Is Cognitive Science?
Cognitive science = study of how humans think — bringing together psychology, neuroscience, linguistics, and anthropology. AI copies useful ideas from it and also tests its theories by building them.
AI: "Can we build intelligence?" — Cognitive science: "How does human intelligence actually work?"
They help each other.
2. Four Key Influences on AI
1. Psychology — Behaviour → Information Processing
- Early psychology was behaviourist (stimulus → response). AI mimicked this with simple reflex agents.
- Later cognitive psychology said: mind does internal reasoning (memory, attention, planning) — AI followed with model-based and goal-based agents that keep internal state and search for plans.
2. Neuroscience — Neurons → Neural Networks
- Brain has ~86B neurons connected together. McCulloch & Pitts (1943) modelled a neuron as a simple threshold unit — direct ancestor of today's perceptron and deep networks.
- Lesson for you: brain inspires architecture (networks, parallel processing) not exact copying.
3. Linguistics — Language → NLP
- Chomsky's work on grammar showed language has structure. AI borrowed grammars, then learned that statistics + data handle real language better.
- This path leads to Module 5's NLP — machine translation, chatbots, etc.
- Beginner insight: Knowing grammar rules ≠ handling messy human speech — need learning from data too.
4. Anthropology & Human Factors — Intelligence in Context
- Intelligence is not just in the head — it uses tools, culture, and environment.
- AI learned to design agents situated in environments (PEAS) and to evaluate AI with human-centred metrics (usability, bias, ethics).
3. How Cognitive Science Changed AI Design
| Human Idea | AI Translation |
|---|---|
| Memory (short/long term) | State, knowledge base, experience buffer |
| Attention (focus on relevant) | Attention mechanism in neural networks, heuristic focus in search |
| Learning by example | Machine learning, reinforcement learning |
| Language understanding | NLP pipelines, transformers |
| Problem solving by mental simulation | State-space search, planning |
4. Simple Analogy to Remember
- Brain : Neuroscience = hardware spec
- Mind : Psychology = software behaviour
- AI tries to write software that shows similar behaviour, sometimes copying hardware style (neural nets), sometimes not (search algorithms).
5. Common Mistake
Thinking AI must copy the brain exactly. It doesn't — airplanes don't flap wings but still fly. AI can be intelligent differently from humans (e.g., A* search has no brain equivalent but solves problems well).
6. One-Line Summary to Write in Exam
Cognitive science showed AI how humans perceive, remember, reason and use language, inspiring agent models, neural networks and NLP while teaching that intelligence depends on both internal processes and the environment.
7. Quick Self-Check
- Name two disciplines inside cognitive science.
- Which AI idea came directly from neuroscience?
- Why is attention relevant to both humans and AI?