logo logo
Home arrow Conscious Machines... arrow Theories of Consciousness arrow Pentti Haikonen's architecture for conscious machines
Saturday, 11 February 2012
 
 
Main Menu
Home
Conscious Machines...
AI Techniques...
Neuroscience...
Researchers...
Publications...
Reviews...
MC Bibliography
Robotics Studio...
Forums...
Blog...
Frontpage
Upcoming Events
22.Feb. 2012

CogSys 2012
Vienna, Austria
04.Apr. 2012

TSC 2012
Tucson, Arizona
17.May. 2012

EAIS 2012
Madrid, Spain
05.Jun. 2012

Cognition & Consciousness
Menorca. Spain
02.Jul. 2012

ASSC 16
Brighton, UK
02.Jul. 2012

Revisiting Turing and his Test
Birmingham, UK
31.Oct. 2012

BICA 2012
Palermo. Italy
Tag Cloud
Associations Books Conferences Conscious Conscious Machines Consciousness Developer Documentación Documentation Español Machine Machine Consciousness Machines Neuroscience Positions Publications Research Researchers Reviews Robotics Robots Spanish Studio VPL
Spotlight
More
Reader's Preferred
MCexperts List
FAQs
Glossary
Site Map
 
Pentti Haikonen's architecture for conscious machines Print E-mail
Written by Trung Doan   
Thursday, 10 December 2009
Article Index
Pentti Haikonen's architecture for conscious machines
Page 2
Page 3
Page 4
Page 5

It can speak, and it has inner speech

If the above utterance pattern is fed to the Haikonen machine's vocalising circuit, the circuit's speaker can produce the sound pattern we recognise as the utterance "ball".

Furthermore, if this pattern is fed to the output side of the aural preprocessor circuit, then the effect is similar to that of the machine hearing the utterance, except in this case it is hearing its own "inner speech".

 

Revisiting perception: How memory helps perception

Even if the Haikonen machine does nothing but gazes at and listens to the world, it continually stores things in its memory and builds up its knowledge of the world.

It has stored its representation of a yellow ball. If it has seen a blue ball, this visual memory would be stored too, together with an episodic memory of when it saw it. If it has seen a ball rolling, this bit of general knowledge would also be stored.

Every time the machine detects a ball, its instantaneous internal representation of this object is broadcast around the brain and this evokes all of its other internal representations to do with balls. Together, they represent the Haikonen machine's knowledge of the world, and evoking them is how the machine perceives.

Compare this: A computer with a camera attached sees a yellow ball. The pixel pattern is processed to detect boundaries, etc., in the end producing sets of numbers representing diameter, size, etc. If this is a quality-control computer on a ball-production line, it checks whether the roundness, diameter, etc., are within specifications, then runs If-Then statements to decide what to do. This is what its 1s and 0s do.

When a young Haikonen machine sees a yellow ball, what the 1s and 0s in its electronics do is to evoke patterns representing "Have seen it before", "Scared of hurling balls", "Have seen a blue ball, too", "Have seen ball rolling", "Have heard the 'ball' utterance". As it accumulates life experience, and makes deductions and decisions, even more representations get created, stored and now evoked: "Have previously accidentally touched it with hand, thus rolling it", "Therefore hand can roll it", etc.

Haikonen calls this perception with "grounded meaning", i.e. the meaning is grounded to the world.

 

It perceives its own body and innards

Assuming that its hands and skin are endowed with touch sensors, then when its hand touches its skin, the concurrence of the 2 sets of signals would allow the machine to deduce that the thing being touched is its skin, rather than something external to it. Likewise, early in the Haikonen machine's cognitive life, as its hand moves in front of its cameras, its camera would, probably like babies discovering their hands and feet, discover that this object does not belong to the world out there but sends the brain sensations from inside.

The battery's preprocessing circuit will also output a signal array, but with far fewer signals. Processing it gives a perception of the machine's own battery level. This then adds to the collection of all the percepts that the electronic brain has about its body, from skin to hands to innards.



Last Updated ( Friday, 11 December 2009 )
 





Lost Password?
No account yet? Register
 Conscious Robots RSS FeedConscious Robots RSS Feed

Find us on Facebook

Follow us on TwitterFollow us on twitter
Spotlight

Machine Consciousness Bibliography Database

 

ConsScale
The Cognitive Machine Consciousness Scale

 
Categories
Last Posts in Forum
 
CR
miel continental