We may not cooperate with friendly machines

Michael Rovatsos

doi:10.1038/s42256-019-0117-1

We may not cooperate with friendly machines

Research output: Contribution to journal › Comment/debate

Abstract

In cooperative games, humans are biased against AI systems even when such systems behave better than our human counterparts. This raises a question: should AI systems ever be allowed to conceal their true nature and lie to us for our own benefit?

Original language	English
Number of pages	2
Journal	Nature Machine Intelligence
DOIs	https://doi.org/10.1038/s42256-019-0117-1
Publication status	Published - 12 Nov 2019

Access to Document

10.1038/s42256-019-0117-1Licence: All Rights Reserved

We may not cooperate_ROVATSOS_DOA12112019_VOR_AllRightsReservedFinal published version, 666 KBLicence: All Rights Reserved

https://www.nature.com/articles/s42256-019-0117-1#rightslinkLicence: All Rights Reserved
https://rdcu.be/bWIVBLicence: All Rights Reserved

Cite this

@article{bc852a50d6474f9e99a94cc0a20aeff0,

title = "We may not cooperate with friendly machines",

abstract = "In cooperative games, humans are biased against AI systems even when such systems behave better than our human counterparts. This raises a question: should AI systems ever be allowed to conceal their true nature and lie to us for our own benefit?",

author = "Michael Rovatsos",

year = "2019",

month = nov,

day = "12",

doi = "10.1038/s42256-019-0117-1",

language = "English",

journal = "Nature Machine Intelligence",

issn = "2522-5839",

publisher = "Springer Nature",

}

TY - JOUR

T1 - We may not cooperate with friendly machines

AU - Rovatsos, Michael

PY - 2019/11/12

Y1 - 2019/11/12

N2 - In cooperative games, humans are biased against AI systems even when such systems behave better than our human counterparts. This raises a question: should AI systems ever be allowed to conceal their true nature and lie to us for our own benefit?

AB - In cooperative games, humans are biased against AI systems even when such systems behave better than our human counterparts. This raises a question: should AI systems ever be allowed to conceal their true nature and lie to us for our own benefit?

U2 - 10.1038/s42256-019-0117-1

DO - 10.1038/s42256-019-0117-1

M3 - Comment/debate

SN - 2522-5839

JO - Nature Machine Intelligence

JF - Nature Machine Intelligence

ER -

We may not cooperate with friendly machines

Abstract

Access to Document

Fingerprint

Cite this