- 2071 Magh · 2+5+1 marks
Knowledge vs learning, acting humanly
Short answer
Knowledge is organized information, facts, and rules a system has for reasoning; learning is the process of acquiring new knowledge or improving performance from experience. Acting humanly means machine behavior indistinguishable from a human, tested by the Turing Test, requiring NLP, knowledge representation, reasoning, learning, and for Total Turing Test, vision and robotics. A well-defined problem has a clear initial state, operators/actions, goal test, and path cost (e.g., 8-puzzle, route finding).
Distinguish between knowledge and learning. What does acting humanly refer to? Explain. Define well defined problems.
Knowledge vs learning
Knowledge is the organised information, facts, rules and relationships that a system already has about a domain and can use for reasoning. Learning is the process of acquiring new knowledge or improving performance from experience, data or instruction.
| Basis | Knowledge | Learning |
|---|---|---|
| Nature | A stored result (what is known) | A process (how it becomes known) |
| Role | Used to solve problems and reason | Used to create or improve knowledge |
| Time | Static at a given moment | Continuous, changes with experience |
| Example | Rule: "IF fever AND cough THEN flu" | A spam filter improving by seeing more emails |
| Representation | Logic, rules, frames, semantic nets | Training of a model (neural net, decision tree) |
Acting humanly
Acting humanly is the AI approach in which a machine's behaviour must be indistinguishable from a human's. It is tested by the Turing Test: if a human interrogator cannot tell whether written responses come from a person or a computer, the computer acts humanly.
To act humanly, a computer needs:
- Natural language processing: to communicate in a human language.
- Knowledge representation: to store what it knows and hears.
- Automated reasoning: to use the stored knowledge to answer questions and draw new conclusions.
- Machine learning: to adapt to new circumstances and detect patterns.
- For the Total Turing Test: computer vision (to perceive objects) and robotics (to manipulate objects and move).
Examples: chatbots, Siri/Alexa and humanoid robots like Sophia.
Well-defined problem
A problem is well defined when it has a clear initial state, a set of operators/actions (successor function), a goal test, and a path cost. Example: 8-puzzle, water jug, route finding.
✓ Answer checked by Claude Opus 5.5 (high effort)
Related questions
- Well defined problem, state-space search · Artificial Intelligence, chapter 2
- Declarative vs procedural knowledge · Artificial Intelligence, chapter 4 · asked 4×
- GA vs fuzzy based learning · Artificial Intelligence, chapter 6
- Analogy vs inductive learning · Artificial Intelligence, chapter 6
- Forces acting on well foundation · Foundation Engineering, chapter 9 · asked 2×
- Knowledge management vs expert systems · Information System, chapter 4
- Supervised vs unsupervised learning · Data Mining, chapter 3