Text Size
Sunday, September 05, 2010
Examples
ConsScale Simplified Rating Examples |
|
| Notes: - Simplified rating provides just an approximation of what could be the real ConsScale level of an implementation. - Implementations (where available) should be rated using the ConsScale Standard Rating process (as described above) in order to obtain a more accurate and realistic measure. - The rating obtained for models which have not yet been fully implemented, will have to be confirmed in the future by the application of the Standard Rating process to the corresponding implementations. - For implementations or models which consider a developmental period, the rating considers the potential final ConsScale that they would achieve at the end of their developmental period. |
| Name of system: | Eliza |
| Description: | One of the first chatterbots (Joseph Weithenbaum, 1966). |
| Architectural Components: | B, Sext, A, R, M (see here for a description) |
| Cognitive Skills: | CS2,1; CS4,1; CS4,2; CS4,3; CS9,3 (see here for a description) |
| Radar Graph: |  |
| CQS Graph: |  |
| CQS Value: | 0.18 |
| ConsScale Level: | LEVEL 2. REACTIVE. |
| Comments: | ELIZA is basically a reactive agent designed to detect and select keywords in the input and, using a script and pattern matching techniques, provide a response in the form of accurate verbal report (CS9,3). Although this agent presents one of the highest level cognitive skills (CS9,3), the final CQS is really low because ConsScale primes a developmental integration of cognitive abilities. In this particular case, it doesn't matter how good is the agent at producing well-formed linguistic reports; if the "mental" content reported is not created by a suitable combination of lower level cognitive abilities, the agent cannot be considered cognitively advanced. |
| Name of implementation: | UT 2004 Adaptive CERA-CRANIUM Bot |
| Description: | Unreal Tournament 2004 autonomous bot implemented using the CERA-CRANIUM cognitive architecture (Raúl Arrabales, 2009). |
| Architectural Components: | B, Sproprio, Sext, A, R, M (see here for a description) |
| Cognitive Skills: | CS2,1; CS3,1; CS3,2; CS4,1; CS4,2; CS4,3; CS4,4; CS4,5; CS4,10; CS5,2; CS5,4 (see here for a description) |
| Radar Graph: |  |
| CQS Graph: |  |
| CQS Value: | 2.69 |
| ConsScale Level: | LEVEL 3. ADAPTIVE. |
| Comments: | Although the implementation complies with some features of levels 4 and 5 it is rated as level 3. ConsScale requires the fulfilment of all cognitive skills of level i and also all lower levels in order to qualify as level i. CQS for a pure level 3 agent is 2.22. The score of this agent (2.69) indicates that some additional features are in place. However, it is far from a level 4 agent who will score 12.21 or more. This can also be noticed in the radar graph, as L4 axis has a short spike (while L2 and L3 axes display full length lines). |
| Name of implementation: | Minimal Architecture for Functional Imagination in CRONOS |
| Description: | Implementation of a functional imagination mechanism that allows an embodied agent to simulate its own actions and their sensory consequences internally, and to extract behavioural benefits from doing so (Hugo Gravato Marques and Owen Holland, 2009). |
| Architectural Components: | B, Sproprio, Sext, A, R, M, Att (see here for a description) |
| Cognitive Skills: | CS2,1; CS3,1; CS3,2; CS4,1; CS4,2; CS4,3; CS4,4; CS4,5; CS4,6; CS4,7; CS4,8; CS4,9; CS4,10; CS7,6 (see here for a description) |
| Radar Graph: |  |
| CQS Graph: |  |
| CQS Value: | 12.24 |
| ConsScale Level: | LEVEL 4. ATTENTIONAL. |
| Comments: | The application of a minimal architecture for functional imagination (CS7,6) to CRONOS is rated as ConsScale level 4 (using Simplified Rating), however being a minimal architecture implementation, the proposal it is very promising in terms of achieving much higher scores (by adding level 5 set shifting capabilities). |
| Name of model: | LIDA Model |
| Description: | The LIDA model is a (not yet fully implemented) comprehensive, conceptual, and computational model of cognition primarily based on the Global Workspace Theory (Franklin et al. 2007; Bernard Baars and Stan Franklin, 2009). |
| Architectural Components: | B, Sproprio, Sext, A, R, M, MN, Att, SsA (see here for a description) |
| Cognitive Skills: | CS2,1; CS3,1; CS3,2; CS4,1; CS4,2; CS4,3; CS4,4; CS4,5; CS4,6; CS4,7; CS4,8; CS4,9; CS4,10; CS5,1; CS5,2; CS5,3; CS5,4; CS5,5; CS6,1; CS6,2; CS6,3; CS6,4; CS6,5; CS7,6 (see here for a description) |
| Radar Graph: |  |
| CQS Graph: |  |
| CQS Value: | 101.36 |
| ConsScale Level: | LEVEL 6. EMOTIONAL. |
| Comments: | As in the case of Haikonen's architecture (below), a comprehensive testing of a full implementation of the LIDA model would be required in order to see if, for instance, the concept of self emerges in the ontology that the agent builds as part of its developmental period. |
| Name of Architecture: | Haikonen's Cognitive Architecture |
| Description: | Cognitive architecture based on distributed signal representations and Haikonen Associative Neurons (Pentti Haikonen, 2007). |
| Architectural Components: | B, Sproprio, Sext, A, R, M, Att, MN, SsA (see here for a description) |
| Cognitive Skills: | CS2,1; CS3,1; CS3,2; CS4,1; CS4,2; CS4,3; CS4,4; CS4,5; CS4,6; CS4,7; CS4,8; CS4,9; CS4,10; CS5,1; CS5,2; CS5,3; CS5,4; CS5,5; CS6,1; CS6,2; CS6,3; CS6,4; CS6,5; CS7,1; CS7,2; CS7,3; CS7,4; CS9,3 (see here for a description) |
| Radar Graph: |  |
| CQS Graph: | |
| CQS Value: | 114.16 |
| ConsScale Level: | LEVEL 6. EMOTIONAL. |
| Comments: | See Doan's article (page 5) on Haikonen's Cognitive Architecture for a discussion on its ConsScale rating. The implementation and testing of a Haikonen Machine would tell us if this architecture is even able to develop the models of self and other selves, thus promoting to ConsScale level 7 (Self-Conscious). |
| < Prev |
|---|
ConsScale - ConsScale Rating
New Scientist. 05 June 2010. Cover Story. pp. 28-31. The topic of mind uploading is addressed in the cover story of this New Scientist issue. ConsScale is quoted as an practical tool to really assess to what extent artificial avatars can be considered as conscious as ourselves. |
www.Conscious-Robots.com is a website about Machine Consciousness and Cognitive Robotics. The Test for Consciousness forum is the place for open discussion about ConsScale and other metrics for consciousness.
Copyright (c) 2010. Raúl Arrabales Moreno. 
This work is licensed under a Creative Commons License.
See related papers and FAQ for more details about ConsScale.
Please send any questions, critiques, or comments to Raúl Arrabales.