Please enable JavaScript.
Coggle requires JavaScript to display documents.
The epistemology of IA - Coggle Diagram
The epistemology of IA
Logical positivism and empiricism
Knowledge
Observable data
Publicly verifiable
Formal languages
Testable through empirical evidence
IA
Structured data
Formal logic, algorithms and statistical validation
Struggles
Meaning
Context
Interpretation
Bias, explainability and validation
AI
Data-dependent
Continuously evolving
Epistemic risk
May be trusted without being justified
Problems
Lack of interpretability
Embedded social bias
Constructivism
Knowledge
Actively constructed
Interaction between the agent and its enviroment
Interpreted through conceptual frameworks
Applied to AI
Based on
Training data
Algorithmic structures
Depends on
How its designed
What can perceive
Challenges for AI
Justification and status
AI modifies knowledge
Lacks human-like understanding
Doubts about AI knowledge and processes correlations