(Sub)Symbolic × (a)symmetric × (non)combinatory: A map of AI approaches spanning symbolic/statistical to neural/ML. Rosenbloom, P. S., Joshi, H., & Ustun, V. In Proceedings of the 7th Annual Conference on Advances in Cognitive Systems, pages 113–131, Cambridge, MA, December, 2019. Cognitive Systems Foundation. Paper abstract bibtex The traditional symbolic versus subsymbolic dichotomy can be decomposed into three more basic dichotomies, to yield a 3D (2×2×2) space in which symbolic/statistical and neural/ML approaches to intelligence appear in opposite corners. Filling in all eight resulting cells then yields a map that spans a number of standard AI approaches plus a few that may be less familiar. Based on this map, four hypotheses are articulated, explored, and evaluated concerning its relevance to both a deeper understanding of the field of AI as a whole and the general capabilities required in complete AI/cognitive systems.
@inproceedings{rosenbloom_subsymbolic_2019,
address = {Cambridge, MA},
title = {({Sub}){Symbolic} × (a)symmetric × (non)combinatory: {A} map of {AI} approaches spanning symbolic/statistical to neural/{ML}},
url = {https://drive.google.com/file/d/1Ynp75A048Mfuh7e3kf_V7hs5kFD7uHsT/view},
abstract = {The traditional symbolic versus subsymbolic dichotomy can be decomposed into three more basic dichotomies, to yield a 3D (2×2×2) space in which symbolic/statistical and neural/ML approaches to intelligence appear in opposite corners. Filling in all eight resulting cells then yields a map that spans a number of standard AI approaches plus a few that may be less familiar. Based on this map, four hypotheses are articulated, explored, and evaluated concerning its relevance to both a deeper understanding of the field of AI as a whole and the general capabilities required in complete AI/cognitive systems.},
booktitle = {Proceedings of the 7th {Annual} {Conference} on {Advances} in {Cognitive} {Systems}},
publisher = {Cognitive Systems Foundation},
author = {Rosenbloom, Paul S. and Joshi, Himanshu and Ustun, Volkan},
month = dec,
year = {2019},
keywords = {UARC, Virtual Humans},
pages = {113--131},
}
Downloads: 0
{"_id":"tn95ivaFc7WBK7rJH","bibbaseid":"rosenbloom-joshi-ustun-subsymbolicasymmetricnoncombinatoryamapofaiapproachesspanningsymbolicstatisticaltoneuralml-2019","author_short":["Rosenbloom, P. S.","Joshi, H.","Ustun, V."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","address":"Cambridge, MA","title":"(Sub)Symbolic × (a)symmetric × (non)combinatory: A map of AI approaches spanning symbolic/statistical to neural/ML","url":"https://drive.google.com/file/d/1Ynp75A048Mfuh7e3kf_V7hs5kFD7uHsT/view","abstract":"The traditional symbolic versus subsymbolic dichotomy can be decomposed into three more basic dichotomies, to yield a 3D (2×2×2) space in which symbolic/statistical and neural/ML approaches to intelligence appear in opposite corners. Filling in all eight resulting cells then yields a map that spans a number of standard AI approaches plus a few that may be less familiar. Based on this map, four hypotheses are articulated, explored, and evaluated concerning its relevance to both a deeper understanding of the field of AI as a whole and the general capabilities required in complete AI/cognitive systems.","booktitle":"Proceedings of the 7th Annual Conference on Advances in Cognitive Systems","publisher":"Cognitive Systems Foundation","author":[{"propositions":[],"lastnames":["Rosenbloom"],"firstnames":["Paul","S."],"suffixes":[]},{"propositions":[],"lastnames":["Joshi"],"firstnames":["Himanshu"],"suffixes":[]},{"propositions":[],"lastnames":["Ustun"],"firstnames":["Volkan"],"suffixes":[]}],"month":"December","year":"2019","keywords":"UARC, Virtual Humans","pages":"113–131","bibtex":"@inproceedings{rosenbloom_subsymbolic_2019,\n\taddress = {Cambridge, MA},\n\ttitle = {({Sub}){Symbolic} × (a)symmetric × (non)combinatory: {A} map of {AI} approaches spanning symbolic/statistical to neural/{ML}},\n\turl = {https://drive.google.com/file/d/1Ynp75A048Mfuh7e3kf_V7hs5kFD7uHsT/view},\n\tabstract = {The traditional symbolic versus subsymbolic dichotomy can be decomposed into three more basic dichotomies, to yield a 3D (2×2×2) space in which symbolic/statistical and neural/ML approaches to intelligence appear in opposite corners. Filling in all eight resulting cells then yields a map that spans a number of standard AI approaches plus a few that may be less familiar. Based on this map, four hypotheses are articulated, explored, and evaluated concerning its relevance to both a deeper understanding of the field of AI as a whole and the general capabilities required in complete AI/cognitive systems.},\n\tbooktitle = {Proceedings of the 7th {Annual} {Conference} on {Advances} in {Cognitive} {Systems}},\n\tpublisher = {Cognitive Systems Foundation},\n\tauthor = {Rosenbloom, Paul S. and Joshi, Himanshu and Ustun, Volkan},\n\tmonth = dec,\n\tyear = {2019},\n\tkeywords = {UARC, Virtual Humans},\n\tpages = {113--131},\n}\n\n","author_short":["Rosenbloom, P. S.","Joshi, H.","Ustun, V."],"key":"rosenbloom_subsymbolic_2019","id":"rosenbloom_subsymbolic_2019","bibbaseid":"rosenbloom-joshi-ustun-subsymbolicasymmetricnoncombinatoryamapofaiapproachesspanningsymbolicstatisticaltoneuralml-2019","role":"author","urls":{"Paper":"https://drive.google.com/file/d/1Ynp75A048Mfuh7e3kf_V7hs5kFD7uHsT/view"},"keyword":["UARC","Virtual Humans"],"metadata":{"authorlinks":{}},"html":""},"bibtype":"inproceedings","biburl":"https://api.zotero.org/users/6976806/collections/GUMH2QKL/items?key=ipCS99jY9KwOteQbpfAW7VKn&format=bibtex&limit=100","dataSources":["Z4B8L2qnYQgDdZhbe","jjKsXjebLR7xyJojc"],"keywords":["uarc","virtual humans"],"search_terms":["sub","symbolic","symmetric","non","combinatory","map","approaches","spanning","symbolic","statistical","neural","rosenbloom","joshi","ustun"],"title":"(Sub)Symbolic × (a)symmetric × (non)combinatory: A map of AI approaches spanning symbolic/statistical to neural/ML","year":2019}