A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind

Edwards, Darren

doi:10.3389/fncom.2024.1395901

Journal article 190 views 18 downloads

A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience...

Darren Edwards

Frontiers in Computational Neuroscience, Volume: 18

Swansea University Author: Darren Edwards

PDF | Version of Record

© 2024 Edwards. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY)
Download (6.5MB)

Check full text

DOI (Published version): 10.3389/fncom.2024.1395901

Abstract

There have been impressive advancements in the field of natural language processing (NLP) in recent years, largely driven by innovations in the development of transformer-based large language models (LLM) that utilize "attention." This approach employs masked self-attention to establish (v...

Full description

Published in:	Frontiers in Computational Neuroscience
ISSN:	1662-5188
Published:	Frontiers Media SA 2024
Online Access:	Check full text
URI:	https://cronfa.swan.ac.uk/Record/cronfa66966
Tags:	Add Tag No Tags, Be the first to tag this record!

first_indexed	2024-07-04T22:32:04Z
last_indexed	2024-07-04T22:32:04Z
id	cronfa66966
recordtype	SURis
fullrecord	<?xml version="1.0" encoding="utf-8"?><rfc1807 xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"><bib-version>v2</bib-version><id>66966</id><entry>2024-07-04</entry><title>A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind</title><swanseaauthors><author><sid>bee507022c083d875238b7802b96cbeb</sid><ORCID>0000-0002-2143-1198</ORCID><firstname>Darren</firstname><surname>Edwards</surname><name>Darren Edwards</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2024-07-04</date><deptcode>HSOC</deptcode><abstract>There have been impressive advancements in the field of natural language processing (NLP) in recent years, largely driven by innovations in the development of transformer-based large language models (LLM) that utilize "attention." This approach employs masked self-attention to establish (via similarly) different positions of tokens (words) within an inputted sequence of tokens to compute the most appropriate response based on its training corpus. However, there is speculation as to whether this approach alone can be scaled up to develop emergent artificial general intelligence (AGI), and whether it can address the alignment of AGI values with human values (called the alignment problem). Some researchers exploring the alignment problem highlight three aspects that AGI (or AI) requires to help resolve this problem: (1) an interpretable values specification; (2) a utility function; and (3) a dynamic contextual account of behavior. Here, a neurosymbolic model is proposed to help resolve these issues of human value alignment in AI, which expands on the transformer-based model for NLP to incorporate symbolic reasoning that may allow AGI to incorporate perspective-taking reasoning (i.e., resolving the need for a dynamic contextual account of behavior through deictics) as defined by a multilevel evolutionary and neurobiological framework into a functional contextual post-Skinnerian model of human language called "Neurobiological and Natural Selection Relational Frame Theory" (N-Frame). It is argued that this approach may also help establish a comprehensible value scheme, a utility function by expanding the expected utility equation of behavioral economics to consider functional contextualism, and even an observer (or witness) centric model for consciousness. Evolution theory, subjective quantum mechanics, and neuroscience are further aimed to help explain consciousness, and possible implementation within an LLM through correspondence to an interface as suggested by N-Frame. This argument is supported by the computational level of hypergraphs, relational density clusters, a conscious quantum level defined by QBism, and real-world applied level (human user feedback). It is argued that this approach could enable AI to achieve consciousness and develop deictic perspective-taking abilities, thereby attaining human-level self-awareness, empathy, and compassion toward others. Importantly, this consciousness hypothesis can be directly tested with a significance of approximately 5-sigma significance (with a 1 in 3.5 million probability that any identified AI-conscious observations in the form of a collapsed wave form are due to chance factors) through double-slit intent-type experimentation and visualization procedures for derived perspective-taking relational frames. Ultimately, this could provide a solution to the alignment problem and contribute to the emergence of a theory of mind (ToM) within AI.</abstract><type>Journal Article</type><journal>Frontiers in Computational Neuroscience</journal><volume>18</volume><journalNumber/><paginationStart/><paginationEnd/><publisher>Frontiers Media SA</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic>1662-5188</issnElectronic><keywords>QBism; consciousness; double slit experiment; functional contextualism; hypergraph; large language model; predictive coding.</keywords><publishedDay>4</publishedDay><publishedMonth>7</publishedMonth><publishedYear>2024</publishedYear><publishedDate>2024-07-04</publishedDate><doi>10.3389/fncom.2024.1395901</doi><url/><notes/><college>COLLEGE NANME</college><department>Health and Social Care School</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>HSOC</DepartmentCode><institution>Swansea University</institution><apcterm>SU College/Department paid the OA fee</apcterm><funders>Swansea University</funders><projectreference/><lastEdited>2024-09-12T14:34:32.7860808</lastEdited><Created>2024-07-04T23:26:55.0036869</Created><path><level id="1">Faculty of Medicine, Health and Life Sciences</level><level id="2">School of Health and Social Care - Public Health</level></path><authors><author><firstname>Darren</firstname><surname>Edwards</surname><orcid>0000-0002-2143-1198</orcid><order>1</order></author></authors><documents><document><filename>66966__31280__87d3332083e343c8ac48fb7e73b36656.pdf</filename><originalFilename>66966.VOR.pdf</originalFilename><uploaded>2024-09-06T15:56:57.8878080</uploaded><type>Output</type><contentLength>6811804</contentLength><contentType>application/pdf</contentType><version>Version of Record</version><cronfaStatus>true</cronfaStatus><documentNotes>© 2024 Edwards. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY)</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>http://creativecommons.org/licenses/by/4.0/</licence></document></documents><OutputDurs/></rfc1807>
spelling	v2 66966 2024-07-04 A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind bee507022c083d875238b7802b96cbeb 0000-0002-2143-1198 Darren Edwards Darren Edwards true false 2024-07-04 HSOC There have been impressive advancements in the field of natural language processing (NLP) in recent years, largely driven by innovations in the development of transformer-based large language models (LLM) that utilize "attention." This approach employs masked self-attention to establish (via similarly) different positions of tokens (words) within an inputted sequence of tokens to compute the most appropriate response based on its training corpus. However, there is speculation as to whether this approach alone can be scaled up to develop emergent artificial general intelligence (AGI), and whether it can address the alignment of AGI values with human values (called the alignment problem). Some researchers exploring the alignment problem highlight three aspects that AGI (or AI) requires to help resolve this problem: (1) an interpretable values specification; (2) a utility function; and (3) a dynamic contextual account of behavior. Here, a neurosymbolic model is proposed to help resolve these issues of human value alignment in AI, which expands on the transformer-based model for NLP to incorporate symbolic reasoning that may allow AGI to incorporate perspective-taking reasoning (i.e., resolving the need for a dynamic contextual account of behavior through deictics) as defined by a multilevel evolutionary and neurobiological framework into a functional contextual post-Skinnerian model of human language called "Neurobiological and Natural Selection Relational Frame Theory" (N-Frame). It is argued that this approach may also help establish a comprehensible value scheme, a utility function by expanding the expected utility equation of behavioral economics to consider functional contextualism, and even an observer (or witness) centric model for consciousness. Evolution theory, subjective quantum mechanics, and neuroscience are further aimed to help explain consciousness, and possible implementation within an LLM through correspondence to an interface as suggested by N-Frame. This argument is supported by the computational level of hypergraphs, relational density clusters, a conscious quantum level defined by QBism, and real-world applied level (human user feedback). It is argued that this approach could enable AI to achieve consciousness and develop deictic perspective-taking abilities, thereby attaining human-level self-awareness, empathy, and compassion toward others. Importantly, this consciousness hypothesis can be directly tested with a significance of approximately 5-sigma significance (with a 1 in 3.5 million probability that any identified AI-conscious observations in the form of a collapsed wave form are due to chance factors) through double-slit intent-type experimentation and visualization procedures for derived perspective-taking relational frames. Ultimately, this could provide a solution to the alignment problem and contribute to the emergence of a theory of mind (ToM) within AI. Journal Article Frontiers in Computational Neuroscience 18 Frontiers Media SA 1662-5188 QBism; consciousness; double slit experiment; functional contextualism; hypergraph; large language model; predictive coding. 4 7 2024 2024-07-04 10.3389/fncom.2024.1395901 COLLEGE NANME Health and Social Care School COLLEGE CODE HSOC Swansea University SU College/Department paid the OA fee Swansea University 2024-09-12T14:34:32.7860808 2024-07-04T23:26:55.0036869 Faculty of Medicine, Health and Life Sciences School of Health and Social Care - Public Health Darren Edwards 0000-0002-2143-1198 1 66966__31280__87d3332083e343c8ac48fb7e73b36656.pdf 66966.VOR.pdf 2024-09-06T15:56:57.8878080 Output 6811804 application/pdf Version of Record true © 2024 Edwards. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) true eng http://creativecommons.org/licenses/by/4.0/
title	A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind
spellingShingle	A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind Darren Edwards
title_short	A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind
title_full	A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind
title_fullStr	A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind
title_full_unstemmed	A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind
title_sort	A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience and LLM architecture for emergent theory of mind
author_id_str_mv	bee507022c083d875238b7802b96cbeb
author_id_fullname_str_mv	bee507022c083d875238b7802b96cbeb_***_Darren Edwards
author	Darren Edwards
author2	Darren Edwards
format	Journal article
container_title	Frontiers in Computational Neuroscience
container_volume	18
publishDate	2024
institution	Swansea University
issn	1662-5188
doi_str_mv	10.3389/fncom.2024.1395901
publisher	Frontiers Media SA
college_str	Faculty of Medicine, Health and Life Sciences
hierarchytype
hierarchy_top_id	facultyofmedicinehealthandlifesciences
hierarchy_top_title	Faculty of Medicine, Health and Life Sciences
hierarchy_parent_id	facultyofmedicinehealthandlifesciences
hierarchy_parent_title	Faculty of Medicine, Health and Life Sciences
department_str	School of Health and Social Care - Public Health{{{_:::_}}}Faculty of Medicine, Health and Life Sciences{{{_:::_}}}School of Health and Social Care - Public Health
document_store_str	1
active_str	0
description	There have been impressive advancements in the field of natural language processing (NLP) in recent years, largely driven by innovations in the development of transformer-based large language models (LLM) that utilize "attention." This approach employs masked self-attention to establish (via similarly) different positions of tokens (words) within an inputted sequence of tokens to compute the most appropriate response based on its training corpus. However, there is speculation as to whether this approach alone can be scaled up to develop emergent artificial general intelligence (AGI), and whether it can address the alignment of AGI values with human values (called the alignment problem). Some researchers exploring the alignment problem highlight three aspects that AGI (or AI) requires to help resolve this problem: (1) an interpretable values specification; (2) a utility function; and (3) a dynamic contextual account of behavior. Here, a neurosymbolic model is proposed to help resolve these issues of human value alignment in AI, which expands on the transformer-based model for NLP to incorporate symbolic reasoning that may allow AGI to incorporate perspective-taking reasoning (i.e., resolving the need for a dynamic contextual account of behavior through deictics) as defined by a multilevel evolutionary and neurobiological framework into a functional contextual post-Skinnerian model of human language called "Neurobiological and Natural Selection Relational Frame Theory" (N-Frame). It is argued that this approach may also help establish a comprehensible value scheme, a utility function by expanding the expected utility equation of behavioral economics to consider functional contextualism, and even an observer (or witness) centric model for consciousness. Evolution theory, subjective quantum mechanics, and neuroscience are further aimed to help explain consciousness, and possible implementation within an LLM through correspondence to an interface as suggested by N-Frame. This argument is supported by the computational level of hypergraphs, relational density clusters, a conscious quantum level defined by QBism, and real-world applied level (human user feedback). It is argued that this approach could enable AI to achieve consciousness and develop deictic perspective-taking abilities, thereby attaining human-level self-awareness, empathy, and compassion toward others. Importantly, this consciousness hypothesis can be directly tested with a significance of approximately 5-sigma significance (with a 1 in 3.5 million probability that any identified AI-conscious observations in the form of a collapsed wave form are due to chance factors) through double-slit intent-type experimentation and visualization procedures for derived perspective-taking relational frames. Ultimately, this could provide a solution to the alignment problem and contribute to the emergence of a theory of mind (ToM) within AI.
published_date	2024-07-04T14:34:32Z
_version_	1809997440984547328
score	11.036706

A functional contextual, observer-centric, quantum mechanical, and neuro-symbolic approach to solving the alignment problem of artificial general intelligence: safe AI through intersecting computational psychological neuroscience...

Similar Items