Foundations of Modern Artificial Intelligence
What Counts as AI?
Definition and scope
Artificial Intelligence (AI) studies agents that perceive, reason, learn, and act under uncertainty to achieve goals. Modern AI spans symbolic reasoning, machine learning, probabilistic inference, optimization, and systems that combine these paradigms.
Core idea
AI is less a single algorithm than a design space of representations, search procedures, learning rules, and decision-making frameworks.
Historical milestones
Selected milestones
| Era | Contribution |
|---|---|
| 1950s | Turing test, early symbolic reasoning |
| 1980s | Expert systems and probabilistic AI |
| 1990s | Statistical learning and graphical models |
| 2010s | Deep learning at scale |
| 2020s | Foundation models and large-scale multimodal systems |
Paradigm shifts
Early AI emphasized explicit rules and search. Statistical AI introduced uncertainty-aware models and generalization from data. Deep learning added hierarchical representation learning, while recent foundation models exploit self-supervised pretraining and large-scale transfer.
Which statement best distinguishes statistical AI from classical symbolic AI?
Statistical AI focuses on probabilistic modeling, inference, and learning from data rather than only on explicit hand-authored rules.
Correct answer: Statistical AI emphasizes learning patterns from data under uncertainty
In one sentence, what is an agent in AI?
This captures the standard AI notion of perception-action loops and goal-directed behavior.
Correct answer: An agent is an entity that perceives its environment and selects actions to achieve goals.
Core mathematical tools
Math toolkit
Why is optimization central to modern AI?
Training typically consists of minimizing a loss or maximizing a likelihood, often with gradient-based methods.
Correct answer: Because many models are trained by minimizing an objective function
Study strategy
When studying AI theory, separate representation, inference, learning, and decision-making. Most advanced systems combine all four.