Difference between revisions of "Main Page"

From aHuman Wiki
Jump to: navigation, search
(Biological Model)
(Biological Model)
 
Line 47: Line 47:
 
* see [[BiologicalLifeResearch]]
 
* see [[BiologicalLifeResearch]]
  
Currently, model is created and has full-coverage status.  
+
Currently, model is successfully created and has full-coverage status.  
 
Data quality is not sufficient to rely on its data.
 
Data quality is not sufficient to rely on its data.
 
To make next step of sustainable quality and scientific nature, model data converted from xml format to relational database (postgres).
 
To make next step of sustainable quality and scientific nature, model data converted from xml format to relational database (postgres).

Latest revision as of 12:59, 10 January 2021

aHuman Project

  • aHuman Project started in 2008 to create alive being by means of a computer
  • After a while it became clear that project belongs to Artificial Life area and to some extent it is attempt to be a God

project.jpg

  • Artificial Life area consists of wet (genetics), hard (robotics) and soft (aHuman case) subdivisions
  • Soft Artificial Life can be achieved using biological architecture or somehow else
  • Any non-biological approach looks to me quite unreliable because it is easy to fall into looking like alive while not being it
  • Biological approach means that we treat human nervous system as mind on brain, where mind is its architecture and brain is wet matter to implement this architecture (like software on hardware)
  • Using biological architecture is hard because mountains of biological data are not adopted to the task and because human mind is still unsolved issue
  • Nevertheless, neurobiology knows almost everything about human mind, with increasing progress last years, and it is hard but reasonable way
  • Finally - aHuman tries to construct creature having the same architectural approach as human but with different sensors and effectors
  • So at least we are not going to be smarter than God

aHuman Platform

  • aHuman project is a platform able to construct and run various targets
  • If succeeded, aHuman can lead to creating (just possible targets):
    • aWee - "personal" (being a person) computer, which is your assistant and friend
    • aYou - make a copy of yourself using Theory of Mind (somehow transhumanistic idea)
    • aPet - make realistic dog or cat to play with
  • Business applications can be industrial control, fair sales and so on
  • Like previous hosting on googlecode (buried, thanks to Google managers), project uses bidirectional svn mapping of site pages (http://usvn.ahuman.org/svn/ahwiki)
  • code is available at:
export from:
svn export http://usvn.ahuman.org/svn/ahengine/trunk_HLAv4/
  • if you would like to do smth in aHuman, please contact me at:
vsavchik@gmail.com
Skype: vladimir.savchik

Biological Model

  • Developments efforts has led to approach of using declarative model and biological model of human, in fact great pass-by-product
  • Biological model is integration of many data of neurobiology and the only source of information, able to present recent data in a consistent manner
  • It worth to mention that that there are some difficulties that make brain model nontrivial and require certain IT experience and interdisciplinary insights
  • Declarative model allows to generate structured information presentations and direct use of biological data by AL-engine
  • Half of this site is created from biological model using specific platform target (http://usvn.ahuman.org/svn/ahengine/trunk_HLAv4/ahuman/target/)
  • see BiologicalLifeResearch

Currently, model is successfully created and has full-coverage status. Data quality is not sufficient to rely on its data. To make next step of sustainable quality and scientific nature, model data converted from xml format to relational database (postgres). Specific software (ahweb) is created to view this information (http://usvn.ahuman.org/svn/ahengine/trunk_HLAv4/ahweb). In HLAv4 ahweb application is modeling tool to view and create model on web, produce xml-model and load into ahengine application, which is brain executable. ahweb application is deployed on web and for readonly access is available anonymously:

see https://ahw.urm.space

Current steps are ahweb development and model improvements:

  • add model modification function to ahweb
  • add model quality model to quantify quality levels by elements with aggregation by hierarchy
  • view unsatisfactory data and prevent from decreasing model quality in modification
  • adjust model data to fit quality model