Show simple item record

dc.contributor.authorKoster, R
dc.contributor.authorBalaguer, J
dc.contributor.authorTacchetti, A
dc.contributor.authorWeinstein, A
dc.contributor.authorZhu, T
dc.contributor.authorHauser, OP
dc.contributor.authorWilliams, D
dc.contributor.authorCampbell-Gillingham, L
dc.contributor.authorThacker, P
dc.contributor.authorBotvinick, M
dc.contributor.authorSummerfield, C
dc.date.accessioned2022-05-11T08:18:02Z
dc.date.issued2022-07-04
dc.date.updated2022-05-10T20:46:23Z
dc.description.abstractBuilding artificial intelligence (AI) that aligns with human values is an unsolved problem. Here, we developed a human-in-the-loop research pipeline called Democratic AI, in which reinforcement learning is used to design a social mechanism that humans prefer by majority. A large group of humans played an online investment game that involved deciding whether to keep a monetary endowment or to share it with others for collective benefit. Shared revenue was returned to players under two different redistribution mechanisms, one designed by the AI and the other by humans. The AI discovered a mechanism that redressed initial wealth imbalance, sanctioned free riders, and successfully won the majority vote. By optimizing for human preferences, Democratic AI offers a proof of concept for value-aligned policy innovationen_GB
dc.identifier.citationPublished online 4 July 2022en_GB
dc.identifier.doi10.1038/s41562-022-01383-x
dc.identifier.urihttp://hdl.handle.net/10871/129585
dc.identifierORCID: 0000-0002-9282-0801 (Hauser, Oliver P)
dc.language.isoenen_GB
dc.publisherNature Researchen_GB
dc.relation.urlhttps://github.com/deepmind/hcmd_daien_GB
dc.rights© The Author(s) 2022. Open Access. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
dc.titleHuman-centred mechanism design with democratic AIen_GB
dc.typeArticleen_GB
dc.date.available2022-05-11T08:18:02Z
dc.identifier.issn2397-3374
dc.descriptionThis is the final version. Available on open access from Nature Research via the DOI in this recorden_GB
dc.descriptionData availability: All human data is available at https://github.com/deepmind/hcmd_daien_GB
dc.descriptionCode availability: Code for reproducing figures is available at https://github.com/deepmind/hcmd_daien_GB
dc.identifier.journalNature Human Behaviouren_GB
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/en_GB
dcterms.dateAccepted2022-05-10
rioxxterms.versionVoRen_GB
rioxxterms.licenseref.startdate2022-05-10
rioxxterms.typeJournal Article/Reviewen_GB
refterms.dateFCD2022-05-10T20:46:28Z
refterms.versionFCDAM
refterms.dateFOA2022-07-06T12:05:19Z
refterms.panelCen_GB


Files in this item

This item appears in the following Collection(s)

Show simple item record

© The Author(s) 2022. Open Access. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
Except where otherwise noted, this item's licence is described as © The Author(s) 2022. Open Access. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.