A Glossary of Terms in the TalaMind Approach
Higher-Level Learning – This term refers collectively to forms of learning required for human-level intelligence, such as learning by creating explanations and testing predictions about new domains based on analogies and metaphors with previously known domains, reasoning about ways to debug and improve behaviors and methods, learning and invention of natural languages and language games, learning or inventing new representations, and in general, self-development of new ways of thinking. The phrase ‘higher-level learning’ is used to distinguish these from lower-level forms of learning investigated in previous research on machine learning (viz. Valiant, 2013).
Higher-Level Mentalities – Used collectively to refer to multi-level reasoning, higher-level learning, understanding (in general and of natural language), imagination, etc.
Human-Level Artificial Intelligence – AI that demonstrates essential capabilities of human-level intelligence, such as human-level generality, originality, natural language understanding, effectiveness and robustness, efficiency, meta-cognition and multi-level reasoning, self-development and higher-level learning, imagination, etc.
Intelligence Kernel – A system of concepts that can create and modify concepts to behave intelligently within an environment. This is a way of describing a ‘baby machine’ approach to human-level artificial intelligence, as a self-extending system of concepts. The term ‘intelligence kernel’ was used by (Jackson, 1979); it has similar meaning to the term ‘seed AI’ used by Yudkowsky (2007).
Multi-Level Reasoning – Used collectively to refer to reasoning capabilities of human-level intelligence at different levels of mentality, such as meta-reasoning, reasoning by analogy, causal and purposive reasoning, abduction, induction, and deduction.
Nested Conceptual Simulation – A Tala agent’s conceptual processing of hypothetical scenarios, with possible branching of scenarios based on alternative events, such as conceptual processing and choices of simulated Tala agents within scenarios.
Tala - The conceptual language defined in Chapter 5 of the thesis, with the proviso that this is only the initial version of the Tala language, open to revision and extension in future work. The name “Tala” is taken from the Indian musical framework for cyclic rhythms, pronounced “Tah-luh”, though the author pronounces it to rhyme with “ballad” and “salad”. The musical term tala is also spelled taal and taala, and coincidentally taal is Dutch for “language”. Tala is also the name of the unit of currency in Samoa.
Tala Agent – A system that implements the TalaMind architecture, to act as an agent within an environment.
TalaMind - Refers to the theoretical approach of the thesis and its hypotheses, and to an architecture the thesis discusses for design of systems according to the hypotheses. TalaMind is also the name of the prototype system illustrating this approach.