Objective and Path
1. The objective of A(G)I
Historical background: is "AI" one problem or many problems?
Major differences between AGI and mainstream AI: to build general-purpose and complete models.
Different understandings (working definitions) of intelligence: focus on structure, behavior, capability, function, or principle?
The advantage of seeing "intelligence" as defined by certain principles: simplicity, elegance, consistency, identity
Working definition: Intelligence is the ability of adapting to the environment while working with insufficient knowledge and resources. An intelligent system should rely on finite processing capacity, work in real time, open to unexpected tasks, and learn from experience.
2. Overall strategy
Different forms of "divide-and-conquer": hybrid, integrated, and unified
The selection of strategy depends on the objective of the research.
The unified approach still allows incremental design and development, and takes "intelligence" as a matter of degree.
3. Formal models
An AGI project should be described on (at least) three levels:
- as a theory of intelligence, in a natural language,
- as a model of the theory, in a symbolic (formal) language,
- as an implementation of the model, in a programming language (plus hardware, if necessary).
Three major traditions of formalization:
dynamical system, computational system, and reasoning system.
The framework of a reasoning system has the following advantages:
- domain independence in state description
- conceptual hierarchy in knowledge representation
- justifiability in step description
- flexibility in process description
4. Types of reasoning system
According to the assumptions about knowledge and resources, three types of inference systems can be distinguished:
- Pure-axiomatic system: In all aspects, the system has sufficient knowledge and resources with respect to the problems to be solved.
- Semi-axiomatic system: In some, but not all, aspects, the system has sufficient knowledge and resources with respect to the problems to be solved.
- Non-axiomatic system: In all aspects, the system has insufficient knowledge and resources with respect to the problems to be solved.
Reading
- Non-Axiomatic Logic: A Model of Intelligent Reasoning, Section 1.1-1.2
- What Do You Mean by "AI"?, Proceedings of AGI-08, 362-373, 2008