|
PresentationsBioinformatics, next generation neuroscience and artificial intelligenceInstitute of Theoretical and Experimental Biophysics of RAS; Institute of Higher Nervous Activity and Neurophysiology of RAS The relationship between bioinformatics, next generation neuroscience and artificial intelligence is presented. Bioinformatics uses traditional artificial intelligence methods to analyze data and in turn provides data and ontologies for AI tools such as medical decision support systems. The concept is introduced: "Next generation neuroscience" - a stage in the development of neurobiology, which is characterized by the use of methods for automatically obtaining and analyzing mass data, the use of non-invasive high-resolution methods, close integration with bioinformatics and developments in the field of traditional AI. Next generation neuroscience uses bioinformatics to obtain and analyze massive genomic data, and also adapts its approaches and methods to analyze big data of its own specific types. Examples of studies that use these data and methods are given, and modern systems for depositing integrated mass data are described. Next generation neuroscience provides data for studying the structural and functional structure and fundamental mechanisms of the functioning of the brain as a carrier of natural intelligence in order to create its functional model capable of performing the main cognitive tasks of a strong artificial intelligence. The concept is introduced: "Average neuromorphic artificial intelligence" - intermediate between highly specialized reactive weak AI and universal proactive strong AI. The average neuromorphic artificial intelligence is characterized by universality, reactivity and limited neuromorphism. Versatility allows it to perform the necessary complex cognitive functions. Reactivity removes the ethical problems of human security and non-increasing the suffering of thinking beings, as well as the task of creating an emotional-motivational block, an attention regulation system, etc. Limited neuromorphism is caused by the incompleteness of our knowledge about the structure of the brain as a substrate of cognitive activity, combined with the lack of need for life support systems, an emotional-motivational block, an attention regulation system, etc. Also, the lack of neuromorphism is caused by the need to solve some problems of input / output, processing and storage of information that are inaccessible to living systems, but desirable for the functioning of artificial intelligence as a tool for human activity. In turn, the creation of an average artificial intelligence will make it possible to systematize the accumulated scientific knowledge at an unprecedented level and fundamentally increase the efficiency of scientific work.
Presentation |
