The Role Of Technology In Food Sustainability – Business Intelligence Info

Wehumansmakesenseoftheworldbylookingforpatterns,filteringthemthroughwhatwethinkwealreadyknow,andmakingdecisionsaccordingly。

Whenwetalkabouthandingdecisionsofftoartificialintelligence(AI),weexpectittodothesame,onlybetter。

Machinelearningdoes,infact,havethepotentialtobeatremendousforceforgood。

Humansarehinderedbyboththeirunconsciousassumptionsandtheirsimpleinabilitytoprocesshugeamountsofinformation。

AI,ontheotherhand,canbetaughttofilterirrelevanciesoutofthedecision-makingprocess,pluckthemostsuitablecandidatesfromahaystackofrésumés,andguideusbasedonwhatitcalculatesisobjectivelybestratherthansimplywhatwe’vedoneinthepast。

Inotherwords,AIhasthepotentialtohelpusavoidbiasinhiring,operations,customerservice,andthebroaderbusinessandsocialcommunities—anddoingsomakesgoodbusinesssense。

Foronething,eventhemostunintentionaldiscriminationcancostacompanysignificantly,inbothmoneyandbrandequity。

Themerefactofhavingtodefendagainstanaccusationofbiascanlingerlongaftertheissueitselfissettled。

Beyondmanagingriskrelatedtolegalandregulatoryissues,though,there’sabroaderargumentfortacklingbias:inarelentlesslycompetitiveandglobaleconomy,noorganizationcanaffordtoshutitselfofffrombroaderinput,morevariedexperiences,awiderrangeoftalent,andlargerpotentialmarkets。

Thatsaid,thealgorithmsthatdriveAIdon’trevealpure,objectivetruthjustbecausethey’remathematical。

HumansmusttellAIwhattheyconsidersuitable,teachitwhichinformationisrelevant,andindicatethattheoutcomestheyconsiderbest—ethically,legally,and,ofcourse,financially—arethosethatarefreefrombias,consciousorotherwise。

That’stheonlywayAIcanhelpuscreatesystemsthatarefair,moreproductive,andultimatelybetterforbothbusinessandthebroadersociety。

Bias:BadforBusiness

WhenpeopletalkaboutAIandmachinelearning,theyusuallymeanalgorithmsthatlearnovertimeastheyprocesslargedatasets。

Organizationsthathavegatheredvastamountsofdatacanusethesealgorithmstoapplysophisticatedmathematicalmodelingtechniquestoseeiftheresultscanpredictfutureoutcomes,suchasfluctuationsinthepriceofmaterialsortrafficflowsaroundaportfacility。

Computersareideallysuitedtoprocessingthesemassivedatavolumestorevealpatternsandinteractionsthatmighthelporganizationsgetaheadoftheircompetitors。

Aswegathermoretypesandsourcesofdatawithwhichtotrainincreasinglycomplexalgorithms,interestinAIwillbecomeevenmoreintense。

UsingAIforautomateddecisionmakingisbecomingmorecommon,atleastforsimpletasks,suchasrecommendingadditionalproductsatthepointofsalebasedonacustomer’scurrentandpastpurchases。

ThehopeisthatAIwillbeabletotakeontheprocessofmakingincreasinglysophisticateddecisions,suchassuggestingentirelynewmarketswhereacompanycouldbeprofitable,orfindingthemostqualifiedcandidatesforjobsbyhelpingHRlookbeyondtheexpecteddemographics。

AsAItakesontheseincreasinglycomplexdecisions,itcanhelpreducebias,consciousorotherwise。

Byexposingabias,algorithmsallowustolessentheimpactofthatbiasonourdecisionsandactions。

Theyenableustomakedecisionsthatreflectobjectivedatainsteadofuntestedassumptions;theyrevealimbalances;andtheyalertpeopletotheircognitiveblindspotssotheycanmakemoreaccurate,unbiaseddecisions。

Imagine,forexample,amajorcompanythatrealizesthatitspasthiringpracticeswerebiasedagainstwomenandthatwouldbenefitfromhavingmorewomeninitsmanagementpipeline。

AIcanhelpthecompanyanalyzeitspastjobpostingsforgender-biasedlanguage,whichmighthavediscouragedsomeapplicants。

Futurepostingscouldbemoregenderneutral,increasingthenumberoffemaleapplicantswhogetpasttheinitialscreenings。

AIcanalsosupportpeopleinmakingless-biaseddecisions。

Forexample,acompanyisconsideringtwocandidatesforaninfluentialmanagementposition:onemanandonewoman。

Thefinalhiringdecisionlieswithahiringmanagerwho,whentheylearnthatthefemalecandidatehasasmallchildathome,assumesthatshewouldpreferapart-timeschedule。

Thatassumptionmaybewellintentioned,butitrunscountertotheoutcomethecompanyislookingfor。

AnAIcouldapplycorrectivepressurebyremindingthehiringmanagerthatallqualificationsbeingequal,thefemalecandidateisanobjectivelygoodchoicewhomeetsthecompany’scriteria。

Thehopeisthatthehiringmanagerwillrealizetheirunfoundedassumptionandremoveitfromtheirdecision-makingprocess。

Atthesametime,bytrackingthepatternofhiringdecisionsthismanagermakes,theAIcouldalertthem—andotherpeopleinHR—thatthecompanystillhassomeremaininghiddenbiasesagainstfemalecandidatestoaddress。

LookforWhereBiasAlreadyExists

Inotherwords,ifwewantAItocountertheeffectsofabiasedworld,wehavetobeginbyacknowledgingthattheworldisbiased。

Andthatstartsinasurprisinglylow-techspot:identifyinganybiasesbakedintoyourownorganization’scurrentprocesses。

Fromthere,youcandeterminehowtoaddressthosebiasesandimproveoutcomes。

TherearemanyscenarioswherehumanscancollaboratewithAItopreventorevenreversebias,saysJasonBaldridge,aformerassociateprofessorofcomputationallinguisticsattheUniversityofTexasatAustinandnowco-founderofPeoplePattern,astartupforpredictivedemographicsusingsocialmediaanalytics。

Inthehighlyregulatedfinancialservicesindustry,forexample,Baldridgesaysbanksarerequiredtoensurethattheiralgorithmicchoicesarenotbasedoninputvariablesthatcorrelatewithprotecteddemographicvariables(likeraceandgender)。

Thebanksalsohavetoprovetoregulatorsthattheirmathematicalmodelsdon’tfocusonpatternsthatdisfavorspecificdemographicgroups,hesays。

What’smore,theyhavetoallowoutsidedatascientiststoassesstheirmodelsforcodeordatathatmighthaveadiscriminatoryeffect。

Asaresult,banksaremoreevenhandedintheirlending。

CodeIsOnlyHuman

Thereasonforthesechecksandbalancesisclear:thealgorithmsthatdriveAIarebuiltbyhumans,andhumanschoosethedatawithwhichtoshapeandtraintheresultingmodels。

Becausehumansarepronetobias,wehavetobecarefulthatweareneithersimplyconfirmingexistingbiasesnorintroducingnewoneswhenwedevelopAImodelsandfeedthemdata。

“Fromtheperspectiveofabusinessleaderwhowantstodotherightthing,it’sadesignquestion,”saysCathyO’Neil,whosebest-sellingbookWeaponsofMathDestructionwaslong-listedforthe2016NationalBookAward。

“Youwouldn’tletyourcompanydesignacarandsenditoutintheworldwithoutknowingwhetherit’ssafe。

Youhavetodesignitwithsafetystandardsinmind,”shesays。

“Bythesametoken,algorithmshavetobedesignedwithfairnessandlegalityinmind,withstandardsthatareunderstandabletoeveryone,fromthebusinessleadertothepeoplebeingscored。

”(TolearnmorefromO’Neilabouttransparencyinalgorithms,readThinkersinthisissue。

)

Don’tDoWhatYou’veAlwaysDone

Toeliminatebias,youmustfirstmakesurethatthedatayou’reusingtotrainthealgorithmisitselffreeofbias,or,rather,thatthealgorithmcanrecognizebiasinthatdataandbringthebiastoahuman’sattention。

SAPhasbeenworkingonaninitiativethattacklesthisissuedirectlybyspottingandcategorizinggenderedterminologyinoldjobpostings。

NothingasovertasNowomenneedapply,whicheveryoneknowsisdiscriminatory,butphraseslikeoutspokenandaggressivelypursuingopportunities,whichareproventoattractmalejobapplicantsandrepelfemaleapplicants,andwordslikecaringandflexible,whichdotheopposite。

Oncehumanscategorizethislanguageandfeeditintoanalgorithm,theAIcanlearntoflagwordsthatimplybiasandsuggestgender-neutralalternatives。

Unfortunately,thisde-biasingprocesscurrentlyrequirestoomuchhumaninterventiontoscaleeasily,butastheamountofavailablede-biaseddatagrows,thiswillbecomefarlessofalimitationindevelopingAIforHR。

Similarly,companiesshouldlookforspecificityinhowtheiralgorithmssearchfornewtalent。

AccordingtoO’Neil,there’snoone-size-fits-alldefinitionofthebestengineer;there’sonlythebestengineerforaparticularroleorprojectataparticulartime。

That’stheneedleinthehaystackthatAIiswellsuitedtofind。

LookBeyondtheObvious

AIcouldbeinvaluableinradicallyreducingdeliberateandunconsciousdiscriminationintheworkplace。

However,themoredatayourcompanyanalyzes,themorelikelyitisthatyouwilldealwithstereotypes,O’Neilsays。

Ifyou’relookingformathprofessors,forexample,andyouloadyourhiringalgorithmwithallthedatayoucanfindaboutmathprofessors,youralgorithmmaygivealowerscoretoablackfemalecandidatelivinginHarlemsimplybecausetherearefewerblackfemalemathematiciansinyourdataset。

ButifthatcandidatehasaPhDinmathfromCornell,andifyou’vetrainedyourAItoprioritizethatcriterion,thealgorithmwillbumpherupthelistofcandidatesratherthansummarilyrulingoutapotentiallyhigh-valuehireonthespuriousbasisofraceandgender。

TofurtherimprovetheoddsthatAIwillbeuseful,companieshavetogobeyondspottingrelationshipsbetweendataandtheoutcomestheycareabout。

Itdoesn’ttakesophisticatedpredictivemodelingtodetermine,forexample,thatwomenaredisproportionatelylikelytojumpoffthecorporateladderatthehalfwaypointbecausethey’restrugglingwithwork/lifebalance。

Manycompaniesfinditalltooeasytoconcludethatwomensimplyaren’tqualifiedformiddlemanagement。

However,acompanycommittedtosmarttalentmanagementwillinsteadaskwhatitisaboutthesepositionsthatmakesthemincompatiblewithwomen’slives。

Itwillthenexplorewhatitcanchangesothatitdoesn’tlosetalentandinstitutionalknowledgethatwillcostthecompanyfarmoretoreplacethantoretain。

Thatcompanymayevenapplyasecondlayerofmachinelearningthatlooksatitsownsuggestionsandmakesfurtherrecommendations:“Itlookslikeyou’retryingtodoX,soconsiderdoingY,”whereXmightbepromotingmorewomen,makingtheworkforcemoreethnicallydiverse,orimprovingretentionstatistics,andYisredefiningjobresponsibilitieswithgreaterflexibility,hostingrecruitingeventsincommunitiesofcolor,orredesigningbenefitspackagesbasedonwhatsimilarcompaniesoffer。

ContextMatters—andContextChanges

EventhoughAIlearns—andmaybebecauseitlearns—itcanneverbeconsidered“setitandforgetit”technology。

Toremainbothaccurateandrelevant,ithastobecontinuallytrainedtoaccountforchangesinthemarket,yourcompany’sneeds,andthedataitself。

Sourcesforlanguageanalysis,forexample,tendtobebiasedtowardstandardAmericanEnglish,soifyou’rebuildingmodelstoanalyzesocialmediapostsorconversationallanguageinput,Baldridgesays,youhavetomakeadeliberateefforttoincludeandcorrectforslangandnonstandarddialects。

StandardEnglishappliesthewordsicktosomeonehavinghealthproblems,butit’salsoapopularslangtermforsomethinggoodorimpressive,whichcouldleadtoanawkwardexperienceifsomeoneconfusesthetwomeanings,tosaytheleast。

Correctingforthat,oraddingmorerulestothealgorithm,suchas“Thewordsickappearsinproximitytopositiveemoji,”takeshumanoversight。

MovingForwardwithAI

Today,AIexcelsatmakingbiaseddataobvious,butthatisn’tthe?

sameaseliminatingit。

It’suptohumanbeingstopayattentionto?

theexistenceofbiasandenlistAItohelpavoidit。

ThatgoesbeyondsimplyimplementingAItoinsistingthatitmeetbenchmarksforpositiveimpact?

Thebusinessbenefitsoftakingthisstepare—orsoonwillbe—obvious。

InIDCFutureScapes’webcast“WorldwideBigData,BusinessAnalytics,andCognitiveSoftware2017Predictions,”researchdirectorDavidSchubmehlpredictedthatby2020perceivedbiasandlackofevidentiarytransparencyincognitive/AIsolutionswillcreateanactivistbacklashmovement,withupto10%ofusersbackingawayfromthetechnology。

However,SchubmehlalsospeculatedthatconsumerandenterpriseusersofmachinelearningwillbefarmorelikelytotrustAI’srecommendationsanddecisionsiftheyunderstandhowthoserecommendationsanddecisionsaremade。

Thatmeansknowingwhatgoesintothealgorithms,howtheyarriveattheirconclusions,andwhethertheydeliverdesiredoutcomesthatarealsolegallyandethicallyfair。

Clearly,organizationsthatcanaddressthisconcernexplicitlywillhaveacompetitiveadvantage,butsimplystatingtheircommitmenttousingAIforgoodmaynotbeenough。

TheyalsomaywishtosupportacademiceffortstoresearchAIandbias,suchastheannualFairness,Accountability,andTransparencyinMachineLearning(FATML)workshop,whichwasheldforthethirdtimeinNovember2016。

O’Neil,whoblogsaboutdatascienceandfoundedtheLedeProgramforDataJournalism,anintensivecertificationprogramatColumbiaUniversity,isgoingonestepfurther。

Sheisattemptingtocreateanentirelynewindustrydedicatedtoauditingandmonitoringalgorithmstoensurethattheynotonlyrevealbiasbutactivelyeliminateit。

Sheproposestheformationofgroupsofdatascientiststhatevaluatesupplychainsforsignsofforcedlabor,connectchildrenatriskofabusewithresourcestosupporttheirfamilies,oralertpeoplethroughasmartphoneappwhentheircreditscoresareusedtoevaluateeligibilityforsomethingotherthanaloan。

AswebegintoentrustAIwithmorecomplexandconsequentialdecisions,organizationsmayalsowanttobeproactiveaboutensuringthattheiralgorithmsdogood—sothattheircompaniescanuseAI?

todowell。

D!

Readmorethoughtprovokingarticlesinthe?

latestissueofthe?

DigitalistMagazine,ExecutiveQuarterly。

AbouttheAuthors:

YvonneBaurisHeadofPredictiveAnalyticsforSapSuccessFactorssolutions。

BrendaReidisVicePresidentofProductManagementforSapSuccessFactorssolutions。

SteveHuntisSeniorVicePresidentofHumanCapitalManagementResearchforSapSuccessFactorssolutions。

FawnFitterisafreelancewriterspecializinginbusinessandtechnology。

Comments

Letsblockads!

(Why?

)

DigitalistMagazine


Tax advice of the week Plant a seed – MoneyWeek

Investments

Sharetips

Stockmarkets

Emergingmarkets

Isas

Pensions

Funds

Bonds

Commodities

Energy

Gold

Bitcoin

CurrencyCorner

Sharedealingaccounts

Alternativefinance

Techstocks



andmore

Economy

Investmentgurus

Entrepreneurssmallbusinesses

TheMoneyWeekportfolioofinvestmenttrusts

Performancetables

FinancialDiary

Spreadbetting

Forextrading

Contractsfordifference(CFDs)

Collectablesandalternativeinvestments

Thebasics

Gettingstartedwithinvesting

Howtoinvest

Financialglossary

OnlineIsaSippproviderscostcomparisontable

Multimedia

TheMoneyWeekPodcast

Merrynsblog

MerrynSomersetWebbinterviews

MoneyWeekvideotutorials

Spreadbettingvideotutorials

FreeMoneyMorningdailyemail

Features

Theworldsgreatestinvestors

Historysworstfinancialcrises

Peopleinthenews

Book,filmandtheatrereviews

Thisdayinhistory

SpendingIt

Housesforsale

Toys

Travel

Wine

Help/About

Aboutus

FAQs

TheMoneyWeekteam

Subscriptions

PrivacyPolicy

CookiePolicy

Policies

AdvertisinginMoneyWeek


r3jnj5qt24l

1WordPress主题大前端DUX6。

0最新版免授权版完美无限制版2传奇复古手游【战神引擎】IP版一键端+安卓端+架设+外网教程3最新游戏(堕落玩偶)[日/英]4。

27G4云赏7。


Toby Edmonds and Justin Stanley reach 100 – 365bet体育盘口_365bet预测_manbetx体育投注

TobyEdmondsandJustinStanleyreach100

July1,20182:21pm。

byAAP

TrainerTobyEdmondshasreachedacenturyofwinnersinaseasonforthefirsttime。

TrainerTobyEdmondsandjockeyJustinStanleyhaveextrareasontocelebratetheirseasonswithbothjoiningthe100winnersclub。

StanleyrodeadoubleatMackayonSaturdaytobecomethe18thjockeyinAustraliatoride100winnersin2017/18。

HeisthirdamongQueenslandridersbehindJeffLloyd(136)andJimByrne(117)

Edmondsofficiallywentpast100whenWinterBridegavehim100-1/2fortheseasonbywinningatCaloundraonSaturdayafterhehadbeenperchedon99-1/2sinceWednesday。

EdmondssonandforemanTrentEdmondssaiditwasarealthrilltogettothecenturyforthefirsttime。

Webeeninthe90sbeforeandsetourselvesagoaltogetto100thisyear,Edmondssaid。


RSS 参考手册

RSS参考手册

RSSchannel元素

请点击元素列中的具体元素,获得更详细的信息。

元素

描述

category

可选的。

为feed定义所属的一个或多个种类。

cloud

可选的。

注册进程,以获得feed更新的立即通知。

copyright

可选。

告知版权资料。

description

必需的。

描述频道。

docs

可选的。

规定指向当前RSS文件所用格式说明的URL。

generator

可选的。

规定用于生成feed的程序。

image

可选的。

在聚合器呈现某个feed时,显示一个图像。

language

可选的。

规定编写feed所用的语言。

lastBuildDate

可选的。

定义feed内容的最后修改日期。

link

必需的。

定义指向频道的超链接。

managingEditor

可选的。

定义feed内容编辑的电子邮件地址。

pubDate

可选的。

为feed的内容定义最后发布日期。

rating

可选的。

feed的PICS级别。

skipDays

可选的。

规定忽略feed更新的天。

skipHours

可选的。

规定忽略feed更新的小时。

textInput

可选的。

规定应当与feed一同显示的文本输入域。

title

必需的。

定义频道的标题。

ttl

可选的。

指定从feed源更新此feed之前,feed可被缓存的分钟数。

webMaster

可选的。

定义此feed的web管理员的电子邮件地址。

RSSitem元素

元素

描述

author

可选的。

规定项目作者的电子邮件地址。

category

可选的。

定义项目所属的一个或多个类别。

comments

可选的。

允许项目连接到有关此项目的注释(文件)。

description

必需的。

描述此项目。

enclosure

可选的。

允许将一个媒体文件导入一个项中。

guid

可选的。

为项目定义一个唯一的标识符。

link

必需的。

定义指向此项目的超链接。

pubDate

可选的。

定义此项目的最后发布日期。

source

可选的。

为此项目指定一个第三方来源。

title

必需的。


M3ETAvMjwP0lZQYSgIqq.jpg

相关推荐2018年度十大爆红营销推广事件盘点美团外卖渠道运营技巧!

为什么用户总是流失?

教你3个阶段用户留存的方法苹果竞价排名专业术语详解及核心数据计算公式!

抖音广告推广:什么样的抖音号值得投放?

青瓜早报:苹果翻新手机印度“合法化”遇阻;小米电视上调3s系列部分产品价格青瓜早报:人民网上线区块链频道;腾讯回应夏总监与米店冲突;阿里90亿美元收购饿了么…运营大叔:为何一夜之间“单身狗属性”刷爆了朋友圈?


The Power Of Habit

原题目:ThePowerOfHabit

关于初入父亲学的每团弄体而言,

我们邑开展了壹段新的人生际涯,

末了尾了壹段新的生活旅途,

而好习惯的养成是必不成微少的,

它能为我们美妙的生活如虎添翼。

ThePowerOfHabit

致无独拥有偶的你

皓天,小酷爱想给父亲家伸荐壹本书,

《习惯的力气》。



揪容







此雕刻本书曾当选FT/高盛2012年度商书简!

荣获2012年亚马逊河编纂铰选

的什父亲滞销书之首!

曾高居2012年《扣儿条约时报》滞销书

排行榜第二名!

习惯不能被免去,却能被顶替。

条需把握“习惯邮路”,念书不清雅察生活中的阴放丢眼色与资深,找到能得到效实感的正确揪容日行为。

学会使用“习惯的力气”,就能让人生与事业洗心革面。







风潮

此雕刻本书己出产版后,收成好评如风潮。

各父亲著名日报和著名干者邑赋予了极好的评价。

ThePowerOfHabit书评

《华尔街日报》华语版:《thepowerofhabit》畅通牒我们:改触动习惯的秘诀是先区别出产“阴放丢眼色”、“揪容日行为”、“奖励”此雕刻叁父亲要斋,然后让它们向更主动的标注的目的转变。

包堕入危急的人也却以用此雕刻种知改触动他们的人生。

《叁联生活周刊》:我们日日行为中40%以上邑是习惯性行为,从驾车到做早米饭和性生活邑成了不需寻求费好多脑儿子的己触动行为。

《thepowerofhabit》壹书畅通牒我们,习惯养成轻善免去难,不外面壹旦你皓白了习惯却以改触动的规律,你就拥有了改触动它们的己在以及责。

ThePowerOfHabit书评

《扣儿条约时报》:什分美不清雅、稀彩绝伦……本书花了好多心血到来论述塑习惯与改触动习惯此雕刻门迷信。

吉姆。

柯林斯(《从优秀到出产色》干者):邑希格的论点万端骈而拥有力:我们应当正眼相看并招认行为的根本驱触动力很难把握,然后才干把此雕刻种根本驱触动力转募化为高效的行为花样。

他拥有着快疾的洞察力,理念既然新鲜又使用。

ThePowerOfHabit书评

老向东方(新正西方教养育科技集儿子团弄实行尽裁剪):《thepowerofhabit》确实是壹本好书,无论关于团弄体习惯的改革,关于企业文皓的优募化,还是关于内阁终止社会办效力动的提升,邑拥有着什分要紧的己创和参考意思。

张道德芬(身眼疾顺手快干家,滞销书《当下的力气》干者):习惯是影响我们一齐生福气的严要紧斋,想要拥有更美妙的人生,我们必须正眼相看己己己的习惯,也坚硬是所谓的揪容性。

本书为我们指出产了壹条皓白的路途,犯得着所拥有人参考、深思。

PS:假设同班们对此雕刻本书感志趣的话,小酷爱更伸荐父亲家看英文版的哦!

英文版的书什分适宜父亲家在当今此雕刻个阶段读呢,在养成好习惯的同时,也兼具提高己己己英语程度的更加处呢。

ThePowerOfHabit

END

往昔日投稿:心思委员会

期望我能壹直陪着你

责编纂:


The Doody Tag Was Invented For Days Like This – The Awl

Believeme,Imorethananyoneunderstandtheimpulsetotitteraboutthisstory,butyoushouldjustremindyourselfthatthisaislegitimatemedicalbreakthroughthatwillsavethelivesofthousandsofpeopleandthatwhileitisonlynaturaltofaceyourdiscomfortwithfecesbyattemptingtodisarmit(thefeces)withhumor,sayingsomethinglike“Thefrontpageoftoday’sNewYorkTimesiscoveredindoody,”isneithermaturenorappropriate,sodon’tdoit。

Anyway,youshouldprintthisoutandreaditlater。

Like,whenyou’reonthetoilet。

Hahaha,getit?