Conclusion - Computational Linguistics and Talking Robots - page 234

flapping wings of birds, but separated in planes into the fixed wing for lift and
the propeller or jet engine for propulsion.
Correspondingly, an artificial DBS agent and its human prototype have in
common that both have a memory connected to external interfaces. An ex-
ample of their many differences, in contrast, is that DBS uses the equality of
core values for sorting proplets into certain token lines of a Word Bank, while
the natural counterpart presumably uses a more differentiated similarity metric
and different principles of storage and retrieval.
Orthogonal to this analogy between artificial flying and artificial cognition
is a crucial difference regarding their interaction with the human user. In an
airplane, the method of being airborne is completely separate from its user-
friendliness for humans. The latter concerns the size of the doors and the seats,
the cabin pressure, the service, etc., while the former concerns the shape of the
wings, the manner of propulsion, the technique of takeoff and landing, etc.
For DBS as a computational theory of cognition, in contrast, maximizing
the similarity between the artificial agent and its natural prototype amounts
directly to maximizing the user-friendliness for humans. 1 This correlation be-
tween the artificial agent and its natural prototype has been preformulated as
the Equation Principle in NLC'06, 1.3.1: 2
12.1.1 T HE E QUATION P RINCIPLE OF D ATABASE S EMANTICS
1. The more realistic the reconstruction of natural cognition, the better the
functioning of the artificial model.
2. The better the functioning of the artificial model, the more realistic the
reconstruction of natural cognition.
The principle is aimed at long-term upscaling. It applies at a level of abstrac-
tion at which it does not matter for completeness of function and data cov-
erage whether cognition is based on natural wetware or electronic hardware. 3
The principle applies to the communication and reasoning aspects 4 of a talking
agent which are (i) concretely observable and (ii) impact directly the quality
of free human-machine communication in natural language.
1 Without giving up any of the applications provided by computers already.
2 The principles presented in Chap. 1 of NLC'06 are (1) the Verification Principle, (2) the Equation
Principle (12.1.1), (3) the Objectivation Principle, (4) the Equivalence Principle for Interfaces (2.6.2),
and (5) the Equivalence Principle for Input/Output (2.6.3).
3 As for “true feelings,” they are equally inaccessible in natural and artificial agents. In fact, they may
be more accessible technically in artificial agents due to their service channel (Chap. 2).
4 It applies less to other aspects of the artificial agent, such as looks - as shown by C3PO and Ripley.