[Fis] Two questions and one remark

[钟义信]

Dear Karl, 


Thank you very much for your very good comments. 


The three points are just my suggestions for the possible focuses of our discussion while "any comments are warmly welcome" as I stated at the end of that e-mail. So, there is no any hard restrictions for our discussion. 


I agree with you that methodology is not independent on, but dependent on, its scientific worldview (see the "declaration"). The change of methodology must imply the change of its scientific worldview. The "three points" did not mention the issue of scientific worldview and this is just for the sake of not involving much too difficult issues ....




Dear Krassimir, The "paradigm for AI" has been described in the "declaration". It can also be expressed in the following way.


The scientific worldview of the paradigm for AI: What is studied in AI is not the matter process, but the information ecological process produced by the subject-object interaction, during which the subject received stimulus, via object information, from an object in the environment and then, for achieving certain goal, the subject should generate (the intelligent) response re-acting on that object. The interaction may need a number of cycles.


This is one of the big "change": from matter issue in the discipline of physical science to the issue of information ecological process in the discipline of AI.


Another big "change" is to change the methodology currently employed in AI study, see below.


The methodology being executed in AI should be the methodology of information ecology, instead of the methodology of mechanical reductionism executed in the discipline of physical science that includes the principles of pure formalism and divide and conquer.


The principles of the methodology of information ecology includes the following: 
(1) the principle of pure formalism is not allowed, but the principle of the trinity of form, utility, and meaning should kept; and 
(2) the principle of "divide and conquer" is not allowed, but the principle of integrity of the information ecological process should be observed.


Dear Krassimir, following the scientific world view and methodology of paradigm for AI, the general (or universal) AI can naturally be produced. 


Due to the rule of 300kb limit, this can only be discussed next time.




Best regards,


 






Prof. Yixin ZHONG
AI School, BUPT
Beijing 100876, China










 

 
 
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listas.unizar.es/pipermail/fis/attachments/20231113/35c6f47c/attachment.html>