Data Storage for artificial intelligence is a crucial issue and indeed quite serious as artificial intelligent systems will be programming themselves. This means that each time they make a new page of data sets perhaps in an XML spreadsheet, to be used as part of the decision matrix for future decisions to determine the next move that it must put an anchor somewhere. That is to say a notation of when and where and why the data sets were created and when to use them, in an appendix type archiving system of its memory storage. Of course when the evaluation process occurs and it must; of each decision made by the artificially intelligent system it must then decide if the data sets imprinted or gathered into its memory are indeed worthy. Should this data be discarded or remain? If it is discarded by perhaps may be some day needed for a future task to prevent the artificially intelligent machine from starting from scratch, it is important to trash can the data sets into an area that is still fully linked by the archival appendix system. Although this is an abstract thought and not necessarily the way in which modern software engineers approach such issues it makes sense to consider this line of reasoning when designing the architectural format for the artificially intelligent self teaching, learning and programming computer system, so do consider this in 2006. |
