Sexuality - PhilArchive

Post on 03-Apr-2023

3 views 0 download

Transcript of Sexuality - PhilArchive

Sexuality

ByJohnDanaher(forthcomingintheOxfordHandbookofEthicsofArtificialIntelligence,editedbyMarkusDubber,FrankPasqualeandSunitDas)Abstract:Sexisanimportantpartofhumanlife.Itisasourceofpleasureandintimacy,andisintegraltomanypeople’sself-identity.ThischapterexaminestheopportunitiesandchallengesposedbytheuseofAIinhowhumansexpressandenacttheirsexualities.Itdoessobyfocusingonthreemainissues.First,itconsiderstheideaofdigisexuality,whichaccordingtoMcArthurandTwist(2017)isthelabelthatshouldbeappliedtothose‘whoseprimarysexualidentitycomesthroughtheuseoftechnology’,particularlythroughtheuseofroboticsandAI.Whileagreeingthatthisphenomenonisworthyofgreaterscrutiny,thechapterquestionswhetheritisnecessaryorsociallydesirabletoseethisasanewformofsexualidentity.Second,itlooksattherolethatAIcanplayinfacilitatinghuman-to-humansexualcontact,focusinginparticularontheuseofself-trackingandpredictiveanalyticsinoptimisingsexualandintimatebehaviour.TherearealreadyanumberofappsandservicesthatpromisetouseAItodothis,buttheyposearangeofethicalrisksthatneedtobeaddressedatbothanindividualandsocietallevel.Finally,itconsiderstheideathatasophisticatedformofAIcouldbeanobjectoflove.Canwebetrulyintimatewithsomethingthathasbeen‘programmed’toloveus?Contrarytothewidely-heldview,thischapterarguesthatthisisindeedpossible.IntroductionInearly2017,theworldborewitnesstoitsfirsthuman-robotmarriage.Zheng

Jiajia,aChineseengineerandAIexpert,hadn’talwaysintendedtomarryarobot.

Hehadspentyearssearchingfora(female)humanpartnerandgrewfrustrated

athislackofsuccess.1Sohedecidedtoputhisengineeringskillstothetestand

1AnotuncommonprobleminChinagivenitsskewedgenderratios.SeeWorldEconomicForum,GlobalGenderGapReport2018,p63,availableathttp://www3.weforum.org/docs/WEF_GGGR_2018.pdf;andalsoViolaZhou‘Chinahasworld’smostskewedsexratioatbirth–again’,SouthChinaMorningPost,27thOctober2016,availableathttps://www.scmp.com/news/china/policies-politics/article/2040544/chinas-

createhisownroboticpartner.Hemarried‘her’inasimple,traditional

ceremonythatwaswitnessedbyhismotherandfriends.2Jiajia’srobotwasn’t

particularlyimpressive.Accordingtothereports,‘she’wasahuman-sizeddoll

withalimitedabilitytorecognizeChinesecharactersandspeaksomebasic

phrases.ButJiajiaplannedtoupgrade‘her’inthenearfuture.

NotlongafterJiajia’snuptials,AikikhoKondo,a35year-oldJapaneseman

livinginTokyo,marriedHatsuneMiku,aholographicvirtualrealitysingerwho

floatsinsideadesktopdevice.3Kondotoofeltunluckyin(human)loveand

plumpedforanartificialpartner.Indoingso,Kondowantedtoberecognisedas

amemberofasexualminorityofpeoplewhoarenotinterestedinhumanlovers.

NeitherJiajianorKondoisalone.Thereisanactiveonlinecommunityof

‘iDollators’whofavourintimacywithartificialdollsoverhumans.Andthereare

nowseveralcompanieseagerlyracingtocreatemoresophisticatedroboticand

artificialcompanions,capableofprovidingtheiruserswithbothsexualintimacy

andemotionalsupport.Weshouldnotbesurprisedbythistrend.Sexand

intimacyareimportantpartsofhumanlifeandtheyhavealwaysbeenmediated

andassistedbytechnology.Sextoysandsexdollscanbefoundgoingback

demographic-time-bomb-still-ticking-worlds-most

2KristinHuang,‘Chineseengineer‘marries’robotafterfailingtofindahumanwife’,SouthChinaMorningPost,4thApril2017,availableathttps://www.scmp.com/news/china/society/article/2084389/chinese-https://www.scmp.com/news/china/society/article/2084389/chinese-engineer-marries-robot-after-failing-find-human-wife3AFP-JIJI,‘Loveinanotherdimension:Japaneseman'marries'HatsuneMikuhologram’,TheJapanTimes,12November2018,availableathttps://www.japantimes.co.jp/news/2018/11/12/national/japanese-man-marries-virtual-reality-singer-hatsune-miku-hologram/#.XFm9vs_7TOQ

thousandsofyearsbackinthearchaeologicalrecord.Thefactthatthelatest

waveoftechnologiesisbeingleveragedtowardsexualendsispartofthislong-

standingtrend.4

Thischapterexaminestheethicalopportunitiesandchallengesposedby

theuseofAIinhowhumansexpressandenacttheirsexualities.Itdoessoby

focusingonthreemainissues.First,itconsidersthequestionofsexualidentity

andasksifweshouldapplyanewsexualidentitylabel–‘digisexuality’–tothose

whoexpressordirecttheirsexualitiestowardsdigital/artificialpartners.5While

agreeingthatthisphenomenonisworthyofgreaterscrutiny,thechapterargues

thatweshouldbeverycautiousaboutrecognisingthisasanewformofsexual

identityasdoingsocanhavestigmatisinganddivisiveeffects.Second,itlooksat

therolethatAIcanplayinfacilitatingandassistinghuman-to-humansexual

intimacy,focusinginparticularontheuseofself-trackingandpredictive

analyticsinoptimisingintimatebehaviour.Itaskswhetherthereissomething

ethicallyobjectionableabouttheuseofsuchAIassistance.Itarguesthatthere

isn’t,thoughthereareethicalrisksthatneedtobeaddressed.Finally,it

considerstheideathatasophisticatedformofAIcouldbeanobjectoflove,

despiteithavingbeen‘programmed’toloveus.Contrarytothewidely-held

view,thischapterarguesthatthisisindeedpossible.

4KateDevlin,TurnedOn:Science,SexandRobots,(London:BloomsburySigma,2018);andHallieLieberman,Buzz:TheStimulatingHistoryoftheSexToy(NewYork:PegasusBooks,2017)5NeilMcArthurandMarkieTwist,‘Theriseofdigisexuality:therapeuticchallengesandpossibilities’(2017)SexandRelationshipTherapy32(3-4):334-344

AIandSexualIdentity

Identityiscentraltohumanexistence.Weallseektodefineandunderstand

ourselvesandothersintermsofdifferentidentitylabels.6Sexualidentitylabels

areanimportantpartofthispatternofclassification.Homosexuality,bisexuality

andhetereosexualityareallnowrecognisedand,forthemostpart,toleratedas

distinctformsofsexualidentity(thoughitwasnotalwaysthus).

Thegeneraltendencytoclassifyourselvesandothersinthismanner

createsatemptationwhenitcomestohowweunderstandthose,likeZheng

JiajiaandAikikhoKondo,whoexpressandenactasexualpreferenceforartificial

partners.Intheirarticle‘Theriseofthedigisexual’,NeilMcArthurandMarkie

Twistsuccumbtothistemptation.7Theyarguethattechnologyplaysan

importantroleinhowpeopleenacttheirsexualdesiresandthatwhenitcomes

tothosewhodisplayamarkedpreferenceforartificialpartners,weshould

recognisethattheyexhibitanewtypeofsexualidentity,namely‘digisexuality’.

Astheyputit:“Manypeoplewillfindthattheirexperienceswiththistechnology

becomeintegraltotheirsexualidentity,andsomewillcometopreferthemto

directsexualinteractionswithhumans.Weproposetolabelthosepeoplewho

considersuchexperiencesessentialtotheirsexualidentity,“digisexuals””.8

McArthurandTwistmakethisargumentwithcircumspectionandcare.

Theypointoutthatsexualorientationsandidentitiesoccuralongacontinuum.

6KwameAnthonyAppiah,TheLiesthatBind:RethinkingIdentity(London:ProfileBooks,2018);andFrancisFukuyamaIdentity:TheDemandforDignityandthePoliticsofResentment(NewYork:Farrar,StrausandGiroux,2018).7McArthurandTwist,n58McArthurandTwist,n5,pp334-335

Somepeoplewilloccasionallyusetechnologytogettheirkicksbutwillretain

strongpreferencesforhuman-to-humancontact.Theysuggestthatonlythose

wholiveprimarilyatoneextremeendofthespectrumdeservethelabel

‘digisexual’.9Theyalsorecognisethatpeoplebelongingtothisgroupwillalmost

certainlysufferfromstigmatisationasaresultoftheirpronouncedsexual

preference,butthencounterthatthissimplyneedstobeunderstoodand

combatted.10Insayingthis,theymakethecaseforusingthe‘digisexuality’label

fromalargelydetached,scientificperspective,suggestingthatthisissomething

thatneedstobeacknowledgedandstudied,notscornedandmaligned.

Iagreethatthereisaphenomenonhereworthyofgreaterscientific

scrutiny,butIthinkweshouldbeverycautiousaboutencouragingthe

widespreaduseofanewsexualidentitylabel,suchas‘digisexuality’,evenfor

suchscientificpurposes.Admittedlythisisnotsomethingthatisnecessarily

underourcontrolsince,aspointedoutabove,weareconstantlyinthebusiness

oflabelingandclassifyingoneanother.Nevertheless,totheextentthatwecan

controlourtendencytolabelandclassifyoneanother,weshouldavoidthe

temptationtorecogniseanewminorityofdigisexuals.Thisstanceisnot

motivatedbyanybigotryordesiretosuppressanewtruthabouthuman

sexuality.Itismotivatedbythedesiretoavoidpathologisingand‘othering’what

shouldbeviewedaspartoftheordinaryrangeofhumansexualdesire.

9McArthurandTwist,n5,p33810McArthurandTwist,n5,p338

Theargumentforthisviewhastwoprongstoit.Thefirstistoclaimthat

therecognitionofaparticularsetofsexualdesiresasadistinctiveidentityor

orientationisnotmetaphysicallymandated.Inotherwords,thereisnothingin

therawdataofhumansexualdesirethatdemandsthatweapplyaparticular

labelorclassificationtothosedesires.Thesecondprongistoarguethattothe

extentthatwedoapplysuchlabels,thereisatendencyforustoignore

importantnuancesintheactualrawdataofhumansexualdesireandforthisto

haveperniciousconsequences.Consequently,sincegroupingsomesetofsexual

desiresintoadistinctiveidentityisnotmetaphysicallymandated,norisit

sociallyorethicallydesirable,weshouldresistthetemptationtodoso.

Let’sexplorebothprongsoftheargumentinmoredetail,startingwith

theclaimthatrecognisinganewsexualidentityisnotmetaphysicallywarranted.

InmakingthisclaimIaminspiredbyatheoryofsexualorientationdevelopedby

SarayAyala:theconceptualacttheoryofsexualorientation.11Thegistofthe

theoryisasfollows.Humanshavemanydifferentphenomenologicalexperiences

intheirlifetimes.Inmanycases,theseexperiencesaremessyandnotfinely

differentiated.Thinkofourauditoryorcolourexperiences.Thoughwedo

perceivedistinctionsbetweendifferentshadesanddifferentmusicalnotes,the

realityofsoundwavesandlightwavesisthattheyblendorfadeintoone

another.Itisonlythroughtheuseofconventionallinguisticlabelsthatwebring

someorderandstructuretothephenomenologicalsoupofexperience.What’s

more,somepeople’sconceptualtoolkitenablesthemtomorefinelydifferentiate

11SarayAyala,‘SexualOrientationandChoice’(2018)JournalofSocialOntology,3(2):249-265.

theirphenomenologicalexperiencesthanothers.Iknowpeoplewhocaneasily

recogniseanddistinguishdifferentnotesandscalesinapieceofmusic.Idonot

havethisability.Ilumptogetherexperiencesthatotherscansplit.

ThepsychologistLisaFeldmanBarretthasarguedthatthissame

phenomenonunderliesouremotionalexperiences.12Theinitial

phenomenologicalrealityofemotionisarawfeelingthatgetsinterpreted

throughaparticularconventionalconceptualtoolkit.Wetranslateourraw

experienceintothefeelingof‘anger’,‘sorrow’or‘joy’(andsoon).Different

culturesparsethephenomenologicalrealityofemotionindifferentways,

groupingandorganisingfeelingsinwaysthatarenotimmediatelyrecognisable

toculturaloutsiders.

Ayalaarguesthatthesameistrueforhowweexperiencesexualdesire.

Overthecourseofalifetime,peoplewillexperiencesexualdesire,arousaland

releaseinresponsetomanydifferentthings.Oftentimesthedesireswillbe

directedatotherpeople,butsometimestheywon’t.Peoplehavebeenknownto

experiencearousalinresponsetoallsortsofenvironmentalstimuli.Whatthen

happensisthatpeoplegrouptheirsexualexperiencestogetherinordertomake

senseoftheirsexualidentitiesandorientations.Indoingthis,someexperiences

areignored,suppressedanddiscounted,whileothersareaccentuated.Youwill

probablydiscountallthosetimesyougotarousedbythevibrationsofthe

schoolbus,butnotthosetimesyougotarousedwhenyoudancedwithyour

12LisaFeldmanBarrett,‘SolvingtheEmotionParadox:CategorizationandtheExperienceofEmotion’(2006)PersonalityandSocialPsychologyReview10(1):20-46

classmateattheschooldance.Youwon’tcallyourselfanautomotive-fetishistno

matterhowmanytimesyougotarousedontheschoolbus.Likewise,and

perhapsmorerealistically,Isuspecttherearemanypeoplewhoprimarilygain

sexualreleasethroughmasturbationandnotthroughintercoursewithanother

humanbeing.Nevertheless,Isuspectthatthemajorityofthosepeopledonot

classifythemselvesasavowedautoeroticists.Theydon’tinterprettheir

masturbatoryexperiencesthroughanidentity-label.Theyseethoseexperiences

asanimportantpartofthefullrangeofdesirablesexualexperiences,allofwhich

arestillbeingactivelypursued.

Thepointhereisthatthesameislikelytobetrueofthosewhogettheir

sexualkicksthroughtechnology,eventhosewhoprimarilydosowithartificial

partners.ConsiderZhouJiajiaandAikhikhoKondo,forexample.Bothofthem

claimtohavesoughtoutartificialpartnersafterfailingtofindloveamongtheir

fellowhumans.Thiswouldsuggestthattheyhaven’tcompletelylostthisformof

sexualdesire.Thedangeristhatifweapply,andencouragethemtoapply,an

identity-labeltotheirnewfoundsexualpreferences,wealsoencouragethemto

discountorsuppresstheotheraspectsoftheirsexualaffect.Theystart

exaggeratingpartofamorediverseanddifferentiatedphenomenologicalreality.

Thisbringsustothesecondprongoftheargument:thatapplyingidentity

labelscanbesociallyandethicallypernicious.Youmightbeprimedtobe

scepticalaboutthis.Youmightpointtootheridentitypoliticalmovementsin

supportofyourscepticismandarguethatowninganidentitylabelcanbeboth

politicallyandpersonallyempowering.Ifyoubelongtoagroupyoufeelless

aloneintheworld.Similarly,ifyouandothermembersofyourgrouparesocially

disadvantaged,bandingtogethercanhelpyoutostandupandagitateforlegal

rightsandprotections.Thishasbeentrueforthefeministmovementandthegay

rightsmovement.Butitisnoteworthythatbothofthesemovementsarosein

responsetopre-existingprejudiceanddiscriminatoryclassification.People

withinthosegroupswerealreadysubjecttoanoppressiveidentity-labelingand

hencesawtheneedtobandtogether,weartheirlabelasamatterofpride,and

workforsocialreform.Intheabsenceofthatpre-existingprejudice,thecasefor

identity-labelingismuchlesspersuasive.Identity-labelingtendstoencourage

divisivenessandothering–the‘us’against‘them’mentality.Peoplequickly

appointthemselvesastheguardiansoftheidentity,creatingcriteriafor

determiningwhobelongsandwhodoesnot.Furthermore,ifbelongingtoa

particularidentitycategorybringswithitcertainsocialbenefitsandlegal

protections,peoplemightbeencouragedtoover-interprettheirexperiencesso

thattheycanfitwithintherelevantgroup:theyforcethemselvesintoagroupso

thattheycanbelong,therebydoingviolencetotheiractualexperience.Inshort,

theidentity-labelingcanfoster,justasoftenasitcancombat,socialdivisionand

polarisation.

Tobeclear,theclaimisnotthatallidentity-labelsareperniciousor

scientificallyinaccurate.Somelabelshavesocialandscientificvalue.Theclaimis

ratherthatidentity-labelshavepowerandshouldbetreatedwithcaution.Sexual

phenomenologyisoftenmorediverseanddifferentiatedthanouridentitylabels

allow.Thismeansthatlumpingsomeoneintoaparticularcategoryisoftennot

warranted.Recognisingandvalorisingtheidentitylabelmayencourageand

incentivisepeopletoforcethemselvestofitintoacategorytowhichtheydonot

belong.So,unlesswearetryingtocombatsomepre-existingsocialprejudiceor

stigmatisation,weshouldveryreluctanttoclassifypeopleas‘digisexuals’.This

doesnotmeanthatwemustignoretherolethatartificialpartnersplayin

people’ssexuallives,orthatwecannotstudythevariousmanifestationsof

‘digisexualities’.Itjustmeansweshouldavoidlabelingpeopleas‘digisexuals’(or

anyothercognatetermlike‘robosexual’).Weshouldacceptthisasjustpartof

thenormalrangeofhumansexualexperience.

AIandSexualAssistance

Sextoysandothersexaidshavelongbeenusedtoassistandcomplement

human-to-humansexualactivity,andAIsandrobotsarealreadywidelyusedto

assistandcomplementnon-sexualhumanactivity.Itshouldbenosurprisethen

tofindAIbeingharnessedtowardsexuallyassistiveends.Wealreadyseesmart

sextoysthattrytolearnfromuserdatatooptimisesexualpleasure;‘quantified

self’appsthatenableuserstotrackandoptimiseaspectsoftheirsexual

performance;andsimpleAIassistantsthathelpwithintimatebehaviour,

includingappsthathelptoautomateorassistwithsendingintimate

communicationstoyourpartner.13DoestheuseofsuchAI-basedsexual

13Fordiscussionsofthedifferentappsandservices,see:DeborahLupton,‘Quantifiedsex:Acriticalanalysisofsexualandreproductiveself-trackingusingapps’,(2015)Culture,HealthandSexuality17(4):440–53;KarenLevy,‘Intimatesurveillance’(2014)IdahoLawReview51:679–93;JohnDanaher,SvenNyholmandBrianEarp‘TheQuantifiedRelationship’(2018)AmericanJournalofBioethics18(2):1-19;JohnDanaher,‘TowardanEthicsofAIAssistants:AnInitialFramework’(2018)PhilosophyandTechnology31(4):629-653;andEvanSelinger‘Today’sAppsareTurningusIntoSociopaths’WIRED26February2014-availableathttps://www.wired.com/2014/02/outsourcing-humanity-apps/;andEvanSelinger,‘Don’toutsourceyourdatingLife’CNN:Edition2May2014-

assistantsraiseanysignificantethicalconcerns?Inpreviouswork,I,alongwith

mycolleaguesSvenNyholmandBrianEarp,analysedeightdifferentethical

concernsonemighthaveabouttheuseofAIinintimaterelationships.14Inthe

interestsofbrevity,Iwilldiscussfourkeyethicalconcernshere:

ThePrivacyConcern:Thisisthebigone.Thisistheconcernthattheuse

ofAIassistantsinintimatesexualrelationshipsconstitutesamajorassaulton

personalprivacy.Thiscouldbebecausepartnersuseservicestospyonone

anotherwithoutconsent.Thisisalreadyaprobleminabusiveintimate

relationships.15ItcouldalsobebecauseAIassistantsareownedandcontrolled

bythirdparties(e.g.companies/corporations)whocapturesexualdatafrom

theirusersandusethistooptimiseandmarkettheirproductsandservices.

Sometimesthisisdonewiththeconsentoftheusers;sometimesitisnot.Indeed,

severallawsuitshavealreadybeensettledbetweencompaniesandusersof

smartsextoysduetothefactthatdatawascollectedfromthosedeviceswithout

theusers’consent.16Ofcourse,violationsofprivacyareageneralconcernwith

digitaltechnology,extendingfarbeyondthesexualorintimateusecase,17but

onemightarguethattheethicalconcernsarehigherinthiscasegiventheunique

importanceofsexualprivacy.

availableathttp://edition.cnn.com/2014/05/01/opinion/selinger-outsourcing-activities/index.html(accessed29/11/2016).14Danaher,NyholmandEarp,n1315Levy,n1316AlexHern,‘VibratormakerorderedtopayoutC$4mfortrackingusers'sexualactivity’,TheGuardian14March2017,availableathttps://www.theguardian.com/technology/2017/mar/14/we-vibe-vibrator-tracking-users-sexual-habits17WoodrowHartzog,Privacy’sBlueprint:TheBattletoControltheDesignofNewTechnologies(Cambridge,MA:HarvardUniversityPress,2018);andShoshanaZuboff,TheAgeofSurveillanceCapitalism(London:ProfileBooks2019)

TheDisengagementConcern:ThisistheconcernthatAIsexual

assistantsmaydistractusfrom,orencourageustodisengagefrom,sexually

intimateactivityandtherebycorrodeorundermineacorepartofthevalueof

thatactivity.Theargumentwouldbethatalotofthegoodofsexualintimacy

(andotherformsofintimacy)stemsfrombeingpresentinthemoment,i.e.

enjoyingthesexualactivityforwhatitis.Butcanyoureallybepresentifyouare

usingsomesex-assistanttotrackthenumberofcaloriesyouburn,orthedecibel

levelreached,orthenumberofthruststhattakeplaceduringsexualactivity?

(These,incidentally,areallrealexamplesofsomeoftheusestowhich

descriptiveandpredictiveanalyticshavebeenputinintimateapps.)18Similarly,

butinanon-sexualcase,EvanSelingerworriesabouttheuseofautomatedand

AI-assistedintimatecommunicationappsonthegroundsthattheycreatethe

impressionthatsomeoneisthinkingaboutandcaringaboutanotherpersonina

particularmomentwheninfacttheyarenotandarelettingtheappdothework

forthem.19Thedisengagementconcernis,onceagain,ageneralconcernabout

digitaltechnology–thinkofallthosecomplaintsaboutthe“anti-social”useof

smartphonesatpartiesandmeetings–butwemightworrythatitisparticularly

problematicintheintimatecasebecauseofhowimportantbeingpresentisto

intimacy.

TheMisdirectionConcern:Relatedtothepreviousconcern,thisisa

concernaboutthekindsofthingsthatAIsexualassistantsmightassistpeople

18Danaher,NyholmandEarp,n13.19Selingern13

with.AIassistantsingeneraltendtoprovideuserswithinformationorprompt

themtodocertainthings.ThesameislikelytooccurwithAIsexualassistants:

theymightgiveusersinformationabouthowtooptimisetheirsexual

experiencesorpromptthemtotryparticularactivities.Oneworryisthatthe

assistantscouldencourageactivitiesthatarenotconducivetogoodsexual

experience.Thisis,indeed,alreadyanexpressedconcernaboutthevarioussex

trackingappsthathavebeencreated.20Asnoted,thoseappsoftenencourage

userstofocusonthingslikethenumberofcaloriesburnedduringsex,the

numberofthrustsduringsex,andthedecibellevelreachedduringsex.One

reasonforthisisthatitisrelativelyeasytotrackandmeasurethesethings.But

thereisnoreasontothinkthatanyofthemiscorrelatedwithgoodsex.Onthe

contrary,focusingonthosemeasuresmightactuallyunderminegoodsex.This

worryisdistinctfromthepreviousonebecauseitisnotabouttheuserbeing

takenoutofthemomentbutratheraboutthemdoingthingsthatarenot

particularlypleasurable/valuableinthemoment.

TheIdeologicalConcern:Afinalconcern,whichisalsorelatedtothe

twoprecedingones,hastodowiththeideologicalimpactofAIsexualassistants

onintimaterelationships.Theconcernisthattheseassistantsmightimposea

certainmodelofwhatanidealintimate/sexualrelationshipisonthepeoplewho

makeuseofthem.Theymight,forexample,recreateandreinforcegender

stereotypesaboutsexualdesireandpreference.KarenLevy,forexample,has

arguedthatmanyintimatetrackingappsreinforcetheviewthatwomenarethe

20OnthiscriticismseeLupton,n13andLevyn13

subjectsofsurveillanceandsexualcontrol.21Othersarguethattheappsmight

encourageaneconomicorexchange-basedmodelofintimaterelationsovera

moreinformal-reciprocationmodel.Thisisbecausethedevicesmightencourage

userstotrackwhodoeswhatforwhomandencouragethemto

optimise/maximisecertainmetrics,alltothedetrimentofwhatatrulyvaluable

intimaterelationshipshouldbe.22

Whatcanbesaidinresponsetotheseconcerns?Well,theprivacyconcern

isprobablythemostserious.IfpartnersuseAIassistantstospyononeanother

ormanipulateoneanother’sbehaviourinanon-transparentway,thenthis

wouldbeamajorworry.Itcouldprovideassistanceandcoverfordominating

andabusiverelationships.Suchrelationshipswillexistintheabsenceof

technologicalassistance,butthetechnologymightmakeiteasiertoimplement

certainformsofdominatingcontrol.Itseemsuncontroversialthentosuggest

thatanyapporservicethatmakesiteasyforoneintimatepartnertospyon

anotherwithouttheother’sconsentshould,ifpossible,bebanned.Spyingby

thirdpartiesshouldalsobelimitedbutistrickiertomanage.Itdoesseemtobe

inherenttodigitaltechnologythatitfacilitatessomekindoftrackingand

surveillance.Wecantrytomitigatetheharmthatisdonebythistrackingand

surveillancethroughrobustlegalprotectionofindividualprivacy.Thislegal

protectionwouldforcethecompaniesthatprovidetherelevantappsand

servicestoputinplacemeasuresthatpreventnon-transparentandnon-

21Levy,n1322Danaher,NyholmandEarp,n13,pp7-8

consensualusesofindividualdata.TheEU’sGeneralDataProtectionRegulation

isastepintherightdirectioninthisregard.

Butitmaywellbethatpeoplearewillingtowaivetheirprivacyrightsin

ordertomakeuseofassistivetechnologies.Thisappearstobethecaseformany

peoplealready.Howmanytimeshaveyouconsentedtodigitalsurveillanceout

ofconvenience?Privacyadvocatescancounterthatthisissimplybecausepeople

donotfullyappreciatethedamagethatcanbedonebythemisuseoftheir

personaldata,butevenstill,formanypeople,convenientaccesstodigital

servicesisoftenfavouredoverprivacy.Thissuggeststhatwhetherornotpeople

arewillingtoforegosomeprivacywhenusingAIsexassistantsmightdependon

whethertheyfindthoseassistantsusefulintheirintimatelives.Iftheydo,then

sexualprivacymightbesignificantlyeroded.

Thisiswheretheotherthreeobjectionscomein.Theyprovidesome

reasontoquestionwhetherAIsexassistantswillinfactbeuseful,highlighting

thevariouswaysinwhichtheymightundermineorcorrodeintimate

relationships.Althougheachofthethreeconcernshassomemerit,theycanbe

overstated.Therearethreereasonsforthis.First,itisimportanttobearinmind

thatthereisnosinglemodelfortheidealintimaterelationship.Different

relationshipmodelsworkfordifferentsetsofpeopleatdifferenttimes.Appsand

assistiveAIthatseemuseless,distractingormisdirectedtosomepeople,might

beuseful,engagingandfulfillingtoothers.Eventheseeminglycomicalexamples

ofsextrackingappsthatgetpeopletoquantifycertainaspectsoftheirsexlife

might,forsomepeople,leadtoamorepleasurableandfulfillingsexlife.Aslong

aspeoplearenotforcedorcompelledtouseparticularAIsexassistants,their

useneednotleadtotheideologicalimpositionofaspecificmodeloftheideal

relationship.Adiversityofappsandassistantscouldprovideroomforpartners

toexploredifferentpossibilitiesinaccordancewiththeirownneedsandwishes.

Second,whilesomeoftheearlyattemptstoprovideAIassistancemightseem

crudeandunsophisticated,theyarelikelytoimproveovertimeandprovide

moreusefulguidance.Thisisbecausethereisreasontothinkthatthetracking

andquantificationmadepossiblebysexandrelationshipappscanbeusedto

goodeffect.Togiveoneexampleofthis,theresearchcarriedoutbytheGottman

Instituteonsuccessfulrelationshipssuggeststhatrelationshipscanbeimproved

ifpartnersexplicitlyrecorddetailsoftheirintimatelives,andfollowcertain

ritualsofconnection.23Theserecommendationsarebasedonextensive,

longitudinalresearchonwhatmakesforasuccessfulintimaterelationship.

Digitalassistantscouldmakeiteasiertoimplementtheserecommendations.

Indeed,theGottmanInstitutealreadyoffersafreesmartphoneappthathelps

couplesimplementsomeofthem.24Onecaneasilyimaginemoresophisticated,

AI-basedversionsofthisappcomingonstreaminthefutureandprovidingfar

moreeffectiveandpersonalisedassistance.Third,totheextentthatworries

remainabouttheeffectofthesetechnologiesonsexualintimacy,theseworries

canbemitigated(toalargeextent)byencouragingmorethoughtfulengagement

withthetechnology.Theproblemsoutlinedaboveareattheirmostsevereif

peopleuseAIassistantsasasubstituteforthinkingforthemselvesandnotasa

complementtothinkingforthemselves.Iftherecouldbeonemajor

23Seehttps://www.gottman.com/24Availableathttps://www.gottman.com/couples/apps/

recommendationmadetothedesignersofAIintimateassistantsitwouldbeto

includeclearwarningstousersthattheservicesandrecommendationsoffered

bytheseassistantsarenotapanaceatoalltheirsexualwoes.Theycanbe

beneficial,butonlyiftheuser(s)criticallyreflectsontheroleoftheservicein

theirownintimatelives.Includingpromptsforsuchcriticalreflectioncouldbea

focusfordesignerswhowishtoencouragetheethicaluseofAIsexassistants.

Thebottomlineis,then,thatalthoughAIassistantscouldundermineand

corrodeourintimateandsexuallives,thereissomereasonforoptimism.The

careful,criticalandnon-dogmaticuseofsuchassistantsmightcomplementand

improveourintimatebehaviour.

AIandLove

Let’scloseoutthischapterbyreturningtothetwomenwhosestoriesI

toldintheintroduction:ZhengJiajiaandAikikhoKondo.Bothofthem‘married’

artificialbeings.Anobviousquestiontoaskiswhattheethicalorphilosophical

statusofthosemarriagesmightbe?Aretheymanifestationsofgenuinelyloving

relationshipsoraretheyslightlyunusualsexualfetishes?Attheoutset,Iwould

emphasisethatanyanswertothisquestionshouldnotbetakentostigmatiseor

shamethosewhoprefersuchrelationships.Butthequestionisworthasking

sinceweattachalotofvaluetolovingrelationshipsandifwecouldhaveloving

relationshipswithAIsandrobots,itmightprovidereasontocreatethem.

Thereis,however,noshortageofoppositiontotheideathatonecouldbe

inalovingrelationshipwitharobot.DylanEvans,forexample,hasarguedthat

thereissomethingparadoxicalabouttheideaofroboticlover.25Hisargument

focusesontheasymmetricalnatureoftherelationshipbetweenahumananda

robot.Presumably,anyroboticloverwillbeprogrammedto‘love’theirhuman

partner.Iftherobotcouldchoosetheirpartnerthenwhatwouldbethepointin

creatingit?Theadvantageofhavingarobotloveroverahumanloveristhefact

thattherobothastoloveyou:thatyouhaveultimatecontroloveritsresponses

toyou(thisdesireforcontrolseemstobeoneofthemotivationsbehindZheng

JiajiaandAikikhoKondo’sactions).Butthiscontrolcomesatacost,accordingto

Evans,becauseacorepartofwhatpeoplewantinalovingrelationshipisa

partner(orpartners)whofreelychoosestobewiththem.Asheputsit,people

wanttheirlover’scommitmenttothemto“bethefruitofanongoingchoice,

ratherthaninflexibleandunreflexivebehaviorpatterns”.26

MichaelHauskelleralsoarguesagainsttheideaofaroboticlover.

Althoughheconcedesthatitmaybepossibletocreatehuman-likerobotsthat

‘appear’tobeinlovewithyou,hecountersthatsuchaloverwouldneverbeas

satisfyingtoyouasahumanlover.FollowingEvans,hearguesthatoneofthe

mainreasonsforthisisthatnomatterhowgoodtheillusionofloveis,there

wouldalwaysbesomereasontosuspectordoubtwhethertherobotreallyloves

you,givenitsorigins.27

25DylanEvans,‘WantingtheImpossible:TheDilemmaattheHeartofIntimateHuman-RobotRelationships’inYorickWilks(ed)CloseEngagementswithArtificialCompanions:KeySocial,Psychological,EthicalandDesignIssues(Philadelphia,PA:JohnBenjaminsPublishingCompany,2010).26Evans,n25,p74-7527MichaelHauskeller,‘AutomaticSweetheartsforTranshumanists’inJohnDanaherandNeilMcArthur(eds)RobotSex:SocialandEthicalImplications(Cambridge,MA:MITPress,2017),p213.

Inamoreextensiveanalysisoftheconceptoflove,SvenNyholmandLily

Frankalsoexpressdoubtsaboutthepossibilityofbeinginalovingrelationship

witharobot.28Exploringdifferentconceptionsofromanticlove(including,the

claimthattobeinloveistobea‘goodmatch’withyourpartner,ortobe

attractedtothe‘distinctiveparticularity’ofyourpartner)theyarguethatwhileit

isnotimpossibletocreatearobotthatmeetstheconditionsneededforaloving

relationship,itwouldbeexceptionallydifficulttodoso,requiringtechnologyfar

inadvanceofwhatiscurrentlyavailable.Inmakingtheircase,theyusethe‘hired

actor’analogytoexpressthebasicproblemwithcreatingaroboticlover:it

seemslikethebestwecanreallydowitharoboticloveristocreateanentity

that‘playsthepart’ofbeinginlovewithyou,butneverquitegraduatesfrom

actingtogenuinelove.29

Thesecriticismsareintuitiveandattractive,buttheyhavesome

problems.Toseewhattheyareitisimportanttodistinguishbetweentwofears

articulatedbythecritics.Thefirst–whichwemightcallthe‘nodepth’fear–is

thatrobotloversareallsurfaceandnodepth.Theyact‘asif’theyloveyoubut

thereisnothingmoretoitthanperformance:theydon’treallyfeelor

consciouslyexperiencetherelevantemotionsthatweassociatewithbeingin

love.Thesecond–whichwemightcallthe‘programming’fear–isthatrobot

loverscannotfreelyandautonomouslychoosetoloveyou.Theywillalwaysbe

28SvenNyholmandLilyEvaFrank,‘FromSexRobotstoLoveRobots:IsMutualLovewithaRobotPossible?’inJohnDanaherandNeilMcArthur(eds)RobotSex:SocialandEthicalImplications(Cambridge,MA:MITPress,2017).29NyholmandFrank,n28,pp223-224.

programmedtoloveyou.Thesetwofearsarerelatedtooneanother–most

allegedrobotloversprobablylackdepthandfreechoice–buttheyarenotthe

samething.Arobotmightbeprogrammedtoloveyouevenifithastheright

kindofexperientialdepthandviceversa.

Arethesetwocriticismsofrobotloversvalid?Let’sconsiderthe‘no

depth’problemfirst.Theeasyrebuttaltothisistosaythatevenifrobots

currentlylacktherequisiteexperientialdepthitispossible,someday,thatthey

willhaveit.Whenthatdayarrives,wecanhaverobotlovers.Themajorproblem

withthisrebuttal,however,isthatitkicksthecandowntheroadandfailsto

grapplewiththephilosophicalissueattheheartofthe‘nodepth’argument,

namely:doesexperientialdepthactuallymatterwhenitcomestodetermining

whetherornotaparticularrelationshipcountsasalovingone?Idon’tthinkit

does.Ifarobotappears,onthesurface,tobeinlovewithyouthenthat’sallit

takesforyoutobeinalovingrelationshipwiththatrobot.

Thismightsoundalittlecrazy,butIdefendthispositiononthegrounds

thatwemust,asapracticalmatter,bebehaviouristswhenitcomesto

understandingtheethicalstatusofourrelationshipswithotherbeings.30In

otherwords,wehavetoapplythemethodologicalbehaviourismofpsychologists

andcomputerscientists(e.g.thebehaviourismattheheartoftheTuringTestfor

machineintelligence)toourethicalrelationshipswithotherbeings.Thecentral

tenetofthis‘ethicalbehaviourism’isthatwhenyoutrytodeterminethemoral

30JohnDanaher,‘ThePhilosophicalCaseforRobotFriendship’(2019)TheJournalofPosthumanStudies2(2)

qualityofyourrelationships(includingyourdutiesandresponsibilities)with

otherbeingsyoucannotuseunobservable,innermentalstatestomakeyour

assessment.Youhavetorelyonexternallyobservablebehaviouraland

functionalpatterns.Youmay,ofcourse,hypothesisetheexistenceofinner

mentalstatestoexplainthoseobservablepatterns.Butanyinferenceyoumake

astothepresenceofthosestatesmustultimatelybegroundedinorguidedbyan

externallyobservablepattern.Theproblemwithmanyofthephilosophical

accountsofwhatittakestobeinalovingrelationshipisthattheyfocus,

implicitlyorexplicitly,onunobservableandinherentlyprivatementalstates

(e.g.feelingsofcommitment/attachment,sincereexpressionsofemotionsand

so).Asaresult,itiseffectivelyimpossibletohaveanyconfidenceintheexistence

oflovingrelationshipsunlessyouacceptthatobservablebehaviouraland

functionalpatternscanprovideepistemicwarrantforourjudgmentsaboutthe

presenceoftherelevantprivatementalstates.Inotherwords,ethical

behaviourismisalready,ofnecessity,theapproachwetaketounderstandingthe

ethicalstatusofourrelationshipswithourfellowhumanbeings.Thismeansthe

‘nodepth’argumentdoesn’twork.Sinceweareunabletoplumbthedepthsof

ourhumanlovers,wecannotapplyadifferentstandardtoroboticlovers.

Thispointhastobefinessedinordertoavoidsomepotentiallyabsurd

interpretations.Forstarters,itisimportanttorealisethatinordertoprovidethe

basisforalovingrelationship,theperformanceand‘illusion’fromarobotwill

needtobeequivalenttotheperformanceandillusionwegetfromahuman

lover.It’sunlikelythatanycurrentlyexistingrobotorAIachievessuch

performativeequivalency.Sothisremains,tosomeextent,afuturepossibility,

notapresentreality.

Similarly,therearesomecounterargumentstoethicalbehaviourismthat

areworthconsidering,ifonlytodeepentheunderstandingofwhatethical

behaviourismentails.Forexample,somepeoplemightarguethatwedorelyon

somethingotherthanbehaviourtodeterminethemoralqualityofour

relationshipswithothers.Perhapsitisbecauseweknowthatourloversare

madeoftherightstuff(biological/organicmaterial)thatweareconfidentthey

canloveus?Orperhapsitisbecauseweknowtheyhavetherightkindof

developmental/evolutionaryhistory?Ifso,thensomeonemightarguethat

robotsandAIwouldstillnotcountas‘proper’loverseveniftheywere

performativelyequivalenttohumanlovers.

Butitishardtoseewhythepresenceorabsenceoftheseotherfactors

shouldhaveshouldhavethateffect.Whatistherationalconnectionbetween

beingmadeoftherightstuff(orhavingtherighthistory)andthecapacityto

formalovingrelationshipwithanother?Supposeyourspousebehavesinaway

thatisentirelyconsistentwiththehypothesisthattheyloveyou.Butthen

suppose,oneday,youlearnthattheyare,infact,analienfromanotherplanet

anddon’tsharethesamebiologicalconstitution.Theycontinuetobehaveasthey

alwaysdid.Shouldyoudoubtwhetheryouareinatrulylovingrelationshipwith

them?It’shardtoseewhy.Therevelationoftheiralienorigins,inandofitself,

shouldnotunderminetheclaimthattheyareinalovingrelationship.The

consistentbehaviouralevidenceofloveshouldtrumptheotherconsiderations.

Thesameshouldholdforaroboticorartificiallover.

Somepeoplemightcomebackandarguethattherearecaseswhereour

faithintheexistenceofalovingrelationshipwouldbeshakenbylearning

somethingabouttheoriginsorhistoryofourhumanlovers.Suppose,for

example,youlearnthatyourhumanloverwas,indeed,ahiredactor,orthatthey

havebeenhavinganaffairforyearswithoutyourawareness.Surelythatwould

undermineyourconfidencethattheyareinalovingrelationshipwithyou?And

surelythatismoreakintowhatitwouldbeliketohavearobotlover?Butthese

counterexamplesdonotwork.Forstarters,itisnotclearthateitherofthese

revelationsshouldshakeourfaithintheexistenceofalovingrelationship.It

seemsplausibletosuggestthatahiredactorcouldgrowtolovethepersonwith

whomtheyhaveaninitiallyfakerelationship,anditalsoseemsplausibleto

suggestthatlovecansurviveinfidelity.Ifthepersonstillbehavesandappearsto

loveyouthenperhapstheydo,despitetheserevelations.Butevenifthat’sa

stretchforsomepeople,Iwouldsuggestthatwhatreallyshakestheirfaithinthe

existenceofalovingrelationshipinbothofthesecasesisthefactthattheywill

acquire(orhavereasontosuspecttheexistenceof)somenewbehavioural

evidencethatcontradictstheoldbehaviouralevidencethatconvincedthemthey

wereinalovingrelationship.Forexample,theymayhavelearned(orstartto

suspect)thattheactorsaysbadthingsaboutthemwhentheyare‘off’thejobor

thattheirpartnerhasbeenplanningtoleavethemforthepersonwiththeyare

havinganaffair.Thisnewbehaviouralevidencemightcompletelyundermine

theirbeliefinalovingrelationshiporatleastpromptthemtoseekoutfurther

behaviouralevidencetoconfirmwhethertheirpartnerstilllovesthem.Either

way,itisbehaviouralevidencethatwilldothedamage(orrepair).Inanyevent,

neitheroftheseexamplesisagoodanalogywiththeroboticlovercase,where

presumablytheroboticnatureandoriginsoftheloverwillbeknownfromday

one.

Whataboutthe‘programming’fear?Evansisrightthatwewant(or,at

least,shouldwant)ourloverstofreelychooseus.Ifarobotisprogrammedor

conditionedtoloveusthenitseemslikethereissomethingsuspiciousorinferior

aboutthekindof‘love’theycangive.Butweshouldn’toverstatethisfeareither.

Itisconceivablethatwecouldcreateroboticloversthatbehave‘asif’theyfreely

chooseus(and,remember,behaving‘asif’theychooseusisenough,following

ethicalbehaviourism).Theroboticlovermightactinficklewayortestitshuman

companion’struecommitment,muchlikeahumanlover.Thiscouldevenbean

attractivequalityinaroboticlover,becauseitmakesitmorelikethehuman-to-

humancase.Thedesireforthisisn’tasbizarreorunfathomableasEvansmakes

out.

Butbeyondthat,thereisalsoreasontodoubtwhetherthepresenceor

absenceof‘programming’shouldundermineourbeliefintheexistenceofa

lovingrelationship.Humansarearguably‘programmed’toloveoneanother.A

combinationofinnatebiologicaldrivesandculturaleducationmakeshumans

primedtofindoneanothersexuallyattractiveandformdeepandlastingbonds

withoneanother.Indeed,peopleoftentalkaboutloveasbeingsomethingother

thanafreeandautonomouschoice.We‘fall’intolove;wedon’tchooseit.We

findourselvesattractedtoothersdespiteourbetterjudgment.Theheartwants

whatitwants,andsoon.Furthermore,insomecultures,arrangedmarriagesand

relationshipsarecommonandwhiletheyseemunusual,maybeevencruel,from

someperspectives,thepartnersinsuchrelationshipsoftengrowtoloveone

anotherandreporthighlevelsofrelationship-satisfaction,sometimeshigherand

oftennoworsethanthesatisfactionofthosein‘autonomous’marriages.31Soitis

notthatunusualtobelievethatlovecanblossomfromsomepre-programming

andpre-arrangingofunions.

Criticsmightdisputetheseexamplesandarguethatthekindof

programminginvolvedinhumanrelationshipsisverydifferentfromthekind

thatwillariseinhuman-robotrelationships.Humansareonlyloosely

programmedtoseekattachment.Theyarenotbrainwashedtoloveaparticular

person.Also,eveninthecaseofarrangedmarriage(wherethereisgreater

restrictionanddirectionofchoice)thepartnersarenotcoercedintothe

relationshiponanongoingbasis.Theycanexercisetheirautonomyafterthe

unionhasformedandescapetherelationshipiftheydesire.

Butitisnotclearthatthedisanalogiesareallthatstrong.Itistruethat,

classically,robotsandAIswereprogrammedfromthetop-downbyparticular

humanprogrammerstofollowhighlyspecifiedinstructions,butthisisnolonger

thenorm.RobotsandAIsarenowprogrammedfromthebottomup,tofollow

31RobertEpstein,MayuriPanditandMansiThakar,‘HowLoveEmergesinArrangedMarriages:TwoCross-culturalStudies’(2013)JournalofComparativeFamilyStudies,44(3):341-360;andPCRegan,SLakhanpal,andCAnguiano,‘RelationshipoutcomesinIndian-Americanlove-basedandarrangedmarriages’(2012)PsychologicalReport110(3):915-24.

learningrules,andtoadapttonewchallengesandcircumstances.Theflexibility

ofthisadaptivelearningisstillratherlimited–weareyettocreateageneralised

formofartificialintelligence–butasthisapproachproliferatesandgrowsthe

allegeddisanalogiesbetweentheprogrammingofhumanloversandrobotlovers

willnarrow.Itwillnolongerbeabsurdtoclaimthatrobotloverscommittous

onthebasisofafreeandongoingchoice,nortoimaginethattheymightfallout

oflovewithusthroughcontinuedlearning.

Noneofthistosaythatpreferringarobotloveroverahumanloverisa

goodthingorthattherearenoethicalproblemswithcreatingrobotlovers.

Thereare.Worriesabouttheobjectificationanddominationofrobotpartners,

aswellasthesocialconsequencesthatthismighthave,havebeenvoicedby

severalcritics.Ihavediscussedtheseworriesatlengthinpreviouswork.32

Similarly,NyholmandFrankarguethatthecreatorsofroboticloversandsexual

partnersmaybeunderanobligationnottomisleadusersastothecapacitiesof

therobotsinquestiontoformlovingrelationships.33Theyworrythat

manufacturersmightbetemptedtoexploittheemotionalvulnerabilityofsome

consumersinordertomaketheirproductsmoreattractive.Whilethisisa

problemwithallconsumerproducts(tosomeextent),itseemslikeaparticularly

acuteproblemforroboticlovers,giventhecentralityandimportanceofsexand

32JohnDanaher,‘RoboticRapeandRoboticChildSexualAbuse:ShouldtheybeCriminalised?’(2017)CriminalLawandPhilosophy11(1):71-95;JohnDanaher‘TheSymbolicConsequencesArgumentintheSexRobotDebate’inDanaherandMcArthur(eds)RobotSex:SocialandEthicalImplications(Cambridge,MA:MITPress,2017);andJohnDanaher,‘RegulatingChildSexRobots:RestrictionorExperimentation?’(2019)MedicalLawReview,forthcoming.33SvenNyholmandLilyEvaFrank‘ItLovesMe,ItLovesMeNot:IsitMorallyProblematictoDesignSexRobotsthatAppearto“Love”TheirOwners?”(2019)Techné,forthcoming

loveinhumanlife.Arelativelystrictsetofrulesmayberequiredtoguard

againstabuse.But,ofcourse,whatisandisnotpermittedbythissetofrules

depends,crucially,onwhatwethinkittakestoformalegitimateloving

relationship.ThisiswhyIhavefocusedonthephilosophicalnatureofloveinthe

precedingdiscussion.IfIamcorrectinmyanalysis,itwillsomedaypossibleto

formalovingrelationshipwitharobotiftherobotcanconvincinglyand

consistentlyperformthepartofbeingalover,andhenceanyrestrictions

imposedtopreventexploitationwillneedtotakethatintoconsideration.

Conclusion

Towrapup,AIandroboticsarebeing,andwillcontinuetobe,usedto

augmentandcomplementhumansexuality.Inthischapter,Ihaveaddressed

threeissuesthatmightariseasaresultandmadethreemainarguments.First,I

havearguedthatweshouldbecautiousaboutrecognisinganewformofsexual

identitythatappliestothosewhoprimarilyexpressandenacttheirsexualities

throughthesetechnologies.Doingsoisnotmetaphysicallymandatedandmay

contributetosocialstigmatisation.Second,IhavearguedthatAIcanbeusedto

assisthumansexualandintimaterelationships.Suchassistanceposesanumber

ofrisks–particularlytoprivacy–buttheserisksshouldnotbeoverstatedand

shouldnotpreventthebeneficialuseofAIsexassistants.Finally,Iarguedthat,

contrarytoanumberofcritics,itispossibletoformalovingrelationshipwitha

robotorAI.It’sabravenewworldintowhichwearestepping.Let’smakeita

sexuallyenrichingone.

Bibliography

JohnDanaherandNeilMcArthur(eds)RobotSex:SocialandEthicalImplications

(Cambridge,MA:MITPress,2017)

JohnDanaher,SvenNyholmandBrianEarp‘TheQuantifiedRelationship’(2018)

AmericanJournalofBioethics18(2):1-19

KateDevlinTurnedOn:Science,SexandRobots(London:Bloomsbury,2018).

MichaelHauskeller,SexandthePosthumanCondition(London:Palgrave

MacMillan,2014)

DavidLevy,LoveandSexwithRobots:TheEvolutionofHuman-Robot

Relationships(NewYork:HarperCollins,2007)

KarenLevy,‘Intimatesurveillance’(2014)IdahoLawReview51:679–93

HallieLieberman,Buzz:TheStimulatingHistoryoftheSexToy(NewYork:

PegasusBooks,2017)

DeborahLupton,‘Quantifiedsex:Acriticalanalysisofsexualandreproductive

self-trackingusingapps’,(2015)Culture,HealthandSexuality17(4):440–53

NeilMcArthurandMarkieTwist,‘Theriseofdigisexuality:therapeutic

challengesandpossibilities’(2017)SexandRelationshipTherapy32(3-4):334-

344

SvenNyholmandLilyEvaFrank,‘FromSexRobotstoLoveRobots:IsMutual

LovewithaRobotPossible?’inJohnDanaherandNeilMcArthur(eds)RobotSex:

SocialandEthicalImplications(Cambridge,MA:MITPress,2017)