《YouGov：2024美國公眾對人工智能（AI）的態度調研報告（英文版）（35頁）.pdf》由會員分享，可在線閱讀，更多相關《YouGov：2024美國公眾對人工智能（AI）的態度調研報告（英文版）（35頁）.pdf（35頁珍藏版）》請在三個皮匠報告上搜索。

1、Public attitudes towards Artificial Intelligence(AI)in the USLiving consumer intelligence| U L Y 2 0 2 4Contents00Overview and segmentation01General perceptions of AI02AI in healthcare03AI in education04AI in law and order05AI in the workplace06Regulation of AI07Methodology08Get in touchPublic attit

2、udes towards artificial intelligence(AI)in the USOverviewThe presence of AI in current affairs and political discussions has increased over the last few years,but how do the public feel towards it and how it might impact their everyday lives in the future?To explore the nuances in public sentiment t

3、hat exist across the US,YouGov PublicYouGov Public conducted a mixed methodsmixed methods study into perceptions towards AI,AI,and specifically its impact in areas the public interact with every day.Our research Our research revealed three dynamic attitudinal segments on AI.The study involved three

4、stages,all conducted in the US in 2023/2024 using the YouGov panel.A parallel study was conducted in the UK.A one-week qualitative online community with 30 participantsA quantitative online online survey of 5,000 adults in the USSegmentation Segmentation analysisanalysisThree Three qualitative quali

5、tative online focus online focus groupsgroupswith22 participantsIntroduction3Introducing the segmentsThe AI AbstainersThe AI OptimistsDiscover living dataArtificial intelligence is increasingly discussed,whether in the press or at home.But how do Americans actually feel about it?And how do they feel

6、 about its use(current and future)in the public sector?To uncover their true perspectives,YouGovs Public team embarked on an in-depth study that segments the US public into three key audiences based on their attitudes,understanding,and use of AI.These dynamic segments provide useful context into the

7、 nuances of public opinion towards and interaction with AI,and how the public can be expected to engage on the topic.The AI IgnorantPublic attitudes towards AI in the US453%5%28%3%HelpsHinders53%42%27%34%40%35%19%27%HealthcareEducationLaw and justiceThe workplace29%Age54%Women46%MenAI helps or hinde

8、rs daily lifeAI IgnorantOverallPositive impact on the future of:AI IgnorantOverall19%23%16%19%23%18-2930-4445-5455-6465+of US adultsPublic attitudes towards AI in the US“When I think of AII think of a robot with a brain”“My knowledge of AI isMinimal.I would say I am only aware in name”.Indicates sig

9、nificant difference against the overall public at 0.95 differenceThe AIIgnorantGeneral perceptions of AIThey have limited knowledge about AI and are less likely to say that AI is important to them.Generally they are indifferent about AI.Use of AI They are not frequent users of AI and are less likely

10、 to read or hear news about it.However,they do not actively try to avoid AI.The future impact of AIThey are more likely to feel neither positively nor negatively about the future use of AI,and they generally think that AI will result in equal amounts of harm and good to society in the future.Regulat

11、ion of AIThey are less likely to knowhow AI ought to be regulated.553%42%27%34%31%16%7%9%HealthcareEducationLaw and justiceThe workplace34%Age60%Women38%MenAI helps or hinders daily lifeAI AbstainersOverallPositive impact on the future of:AI AbstainersOverallof US adultsPublic attitudes towards AI i

12、n the US10%18%18%22%31%18-2930-4445-5455-6465+Indicates significant difference against the overall public at 0.95 differenceThe AIAbstainersGeneral perceptions of AIThey have limited knowledge of AI but generally feel very negatively towards it.They are more likely to believe that it makes fraudulen

13、t activity likely.Use of AI They are not frequent users of AI and are more likely to say that they have no interaction with AI in their daily life.The future impact of AIThey feel negatively about the future of AI,do not think that we should embrace AI or that it will help speed up innovation.Regula

14、tion of AIThey think that without the correct regulation,AI could get out of control.They are less likely to think AI should continue to be developed.“AI means loss of jobs and further encroachment on personal privacy“I dont use anything AI because I want to do things on my own and use my own brain”

15、653%5%11%21%HelpsHinders53%5%74%1%HelpsHindersAI OptimistsOverall53%42%27%34%84%72%52%63%HealthcareEducationLaw and justiceThe workplaceAI OptimistsOverall37%Age39%Women60%MenAI helps or hinders daily lifePositive impact on the future of:of US adultsPublic attitudes towards AI in the USIndicates sig

16、nificant difference against the overall public at 0.95 difference33%30%13%11%13%18-2930-4445-5455-6465+The AIOptimistsGeneral perceptions of AIAI is important to them,and they are more likely to know about it and its uses.They generally believe that AI helps society overall,and that it will continue

17、 to do so.Use of AI They are frequent users of AI,and actively look to use it in their daily lives.The future impact of AIThey feel positively about the future of AI and its impact on various work sectors.They believe that we should embrace AI.Regulation of AIThey are more likely to think that AI sh

18、ould be allowed to continue to be developed,but with some regulations in place.“AI has great potential for good as it can sift through data faster and deeper than humans-already showing potential in medical field“AI is technology that could be used to assist you with tasks and productivity which cou

19、ld possibly grow and learn to be more efficient over time”7General perceptions of AITalk to a researcherDevelopments in AI is not currently considered a top global issue General perceptions of AI9Only 18%of the US public rank developments in AI and technology as one of the most important global issu

20、es.And of the 9 global issues tested,only access to education is considered less important.Issues such as war and terrorism(58%),climate change and the environment(54%),and the economy and employment(36%)are far more likely to be ranked as a top 3 global issue.AI Optimists AI Optimists are more like

21、ly to rank developments in AI as a top 3 global issue,with 27%doing so,while just 15%of AI AbstainersAI Abstainers and 8%of AI Ignorant AI Ignorant Americans did so.%Ranked as top 3 global issueWar and terrorism58%The economy and employment54%Climate change and the environment36%Food and water secur

22、ity33%Immigration and refugees32%Global health(diseases and pandemics)29%Inequality18%Developments in AI and technologyDevelopments in AI and technology18%Access to education11%Looking at the following global issues,please rank which you think are the most important at this time.Please rank up to th

23、ree.Base:All(n=5,074)General perceptions of AI1031%21%31%18%OverallDrawbacks outweigh benefitsAbout equalBenefits outweigh drawbacksDont knowWhen asked about the AIs impact on wider society,the same proportion of the public think that the benefits outweigh the drawbacks as think the opposite(31%).Mo

24、st AI Optimists AI Optimists think that the benefits of AI outweigh the drawbacks(61%),compared to the majority of AI Abstainers,AI Abstainers,who think that the drawbacks outweigh the benefits(70%).AI Ignorant AI Ignorant are more likely than the public overall to say the impact of AI is about equa

25、l(26%)or they dont know(39%).Impact of AI on wider societyThe public are divided on whether the benefits of AI outweigh the drawbacksIn the qualitative groups,AI Ignorant AI Ignorant and OptimistsOptimists were more positive about AIs current impact on society.However,all segments expressed similar

26、concerns about the negative impacts of AI.Participants point to recent stories of AI being used to spread election disinformation through deepfakes as evidence of AIs potential to cause serious harm.“AI is being used to influence elections so that we now have truly ignorant and uneducated people in

27、positions of power.”-AI Ignorant“I mean,were all nervous about AI-generated robocalls and images around the election.”-AI OptimistsDespite this,AI IgnorantAI Ignorant and OptimistsOptimists are still cautiously hopeful about the impact that AI will have on society in the future.Thinking about artifi

28、cial intelligence(AI),in general,which of the following comes closest to your viewpoint?Base:All(n=5,074)Only a small minority of those who use AI think that it hinders their daily lifeGeneral perceptions of AIWhen asked about its impact on their daily lives,a majority of those who use AI feel that

29、it helps them(54%),while just 5%think that it hinders them in their daily life.Over a third(37%)think it neither helps nor hinders them.AI Optimists AI Optimists are more likely to think that AI currently helps them in their daily life(74%).By comparison,28%of the AI Ignorant AI Ignorant who interac

30、t with AI and 11%of AI Abstainers AI Abstainers think that AI currently helps their daily lives.A fifth of AI Abstainers AI Abstainers(21%)think that their current use of AI hinders them.5%37%54%4%OverallHindersNeither helps nor hindersHelpsDont knowImpact of AI on users daily livesIn the qualitativ

31、e groups,there was no consensus as to what constitutes“AI,”and many of the AI AI Ignorant Ignorant and AI Abstainers AI Abstainers felt that AI has hindered them.Care and service-oriented processes that were once managed by humans and are becoming more automated over time are the most heavily-cited

32、source of AI making their daily lives harder.“I called an airline the other day,and they said I could hang up and have the conversation by text and when they text me,it was only a bot!”-AI Ignorant“Customer service has kind of imploded because of AI.”-AI Abstainers11In general,would you say artifici

33、al intelligence(AI)helps or hinders you in your life,or does it neither help nor hinder?Base:All who use AI(n=2,614)There are higher levels of distrust in how AI might be used in the future than for how it is currently used Trust/Distrust in how AI is currently used/might be used in the futureGenera

34、l perceptions of AIJust under three in ten(28%)of the public trust how AI is currently used,with four in ten(39%)expressing distrust.A similar number(26%)trust how AI might be used in the future,though almost half(48%)distrust its future uses.AI Abstainers AI Abstainers have the highest levels of di

35、strust in how AI is currently used(75%),which increases to 85%for future uses.The AI Ignorant AI Ignorant are most likely to neither trust nor distrust how AI is currently used(48%).Around a fifth(22%)distrust how it is currently used and 16%trust it.A majority of AI Optimists AI Optimists trust how

36、 AI is currently used(57%)with slightly fewer expressing trust in how it might be used in the future(51%).While less likely to distrust AIs future uses than other segments,30%of this group still hold this view.28%39%26%48%Net:TrustNet:TrustNet:DistrustNet:DistrustCurrently usedCurrently usedFuture u

37、seFuture use12Thinking generally,how much trust or distrust do you have in.Base:All(n=5,074)The public are most likely to think AIwill positively impact healthcare General perceptions of AIHalf(53%)think that AI will have a positive impact on healthcarehealthcare in the future,which makes it the sec

38、tor the public feel most positively about when it comes to AI.Even 31%of AI Abstainers AI Abstainers think that AI could have a positive impact on healthcare(although this is notably less than the 84%of AI Optimists AI Optimists who think so).Four in ten(42%)think that AI will have a positive impact

39、 on educationeducation.Seven in ten(72%)AI AI Optimists Optimists think AI will have a positive impact on education,while just 35%of AI Ignorant AI Ignorant and 16%of AI Abstainers AI Abstainers do.Opinion surrounding AI and the workplaceworkplace is the most divided.A third(34%)of the public think

40、AI will have a positive impact on the workplace,while a similar proportion(35%)think that its impact will be negative.This is most strongly felt by AI Abstainers AI Abstainers(68%of which think that AI will negatively impact the workplace).The public are also split on what impact AI will have on law

41、 and justicelaw and justice.Indeed,27%think it will have a positive impact,compared to 33%who think its impact will be negative and 29%who think it will have neither a positive nor negative impact.This is the sector that AI Optimists AI Optimists are the least positive about,with 27%saying the impac

42、t will be neither positive nor negative and 16%saying the impact will be negative.Nonetheless,half of this segment(52%)still think AI will have a positive impact.33%35%28%17%27%34%42%53%Law and justiceLaw and justiceThe workplaceThe workplaceEducationEducationHealthcareHealthcarePositiveNegativeImpa

43、ct of AI on each of the following13In general,how positive or negative an impact do you think artificial intelligence(AI)will have on each of the following in the future?Or will it have no impact?Base:All(n=5,074)AI in healthcareTalk to a researcherA majority think its acceptable for AI to manage an

44、d book appointments,but the public are divided on other aspectsof healthcareAI in healthcare1513%31%32%47%46%52%64%45%37%29%28%24%To manage and bookappointmentsTo review scans anddiagnose illnessesTo allocate medicalresources in hospitalsFor deciding whatmedication to prescribeTo make decisions abou

45、thealth insurance coverageFor an initial primary careconsultationNet:AcceptableNet:UnacceptableAlmost two thirds(64%)think its acceptable for AI to manage and book healthcare appointments,with even 51%of AI Abstainers AI Abstainers and AI IgnorantAI Ignorant holding this view.The public are also mor

46、e likely to think its acceptable for AI to be used to review scans and diagnose illnesses(45%),than not(31%).When it comes to allocating medical resources in hospitals the public are split(37%acceptable,32%unacceptable)while leaning towards it being unacceptable for AI to be involved in other decisi

47、on-making processes such as deciding prescriptions and health insurance coverage.Just 24%think it is acceptable for AI to perform an initial primary care consultation,compared to 52%who are against this.AI Optimists AI Optimists are open to AI being used for this purpose though with 46%thinking this

48、 is acceptable,38%against.Acceptable/Unacceptable for AI to doPlease say whether you think it would be acceptable or unacceptable to use artificial intelligence(AI)in each of the following aspects of healthcare:Base:All(n=5,074)For an initial primary care consultationTo make decisions about health i

49、nsurance coverageFor deciding what medication to prescribeTo allocate medical resources in hospitalsTo review scans and diagnose illnessesTo manage and book appointments16Just over a fifth(23%)say that they would prefer AI to manage and book their healthcare appointments,with a further 39%saying the

50、y wouldnt mind whether AI or a human perform this task,so long as it was done well.For the remaining aspects of healthcare tested,a majority of the public would prefer a human over AI.Most notably,seven in ten would prefer a human to decide what medication to prescribe them(70%)and for an initial pr

51、imary care consultation(72%).Even a majority of AI OptimistsAI Optimists would prefer a human to do each of these tasks(60%and 62%respectively).The public would prefer a human to perform all healthcare tasks,except to manage and book appointmentsIn the qualitative groups,healthcare was the only fiel

52、d that all segments agreed that AI could have a positive impact on,though AbstainersAbstainers were more cautious.They want AI to be used as a tool to help healthcare workers do their jobs more efficiently,but do not want AI to replace human interaction in healthcare.“I like the idea of some self-di

53、agnosing of medical issues.with the overburdened medical system we have today,it can save precious time.”-AI IgnorantAI could maybe help with medical diagnoses”-AI Abstainers23%12%11%10%6%6%32%55%58%61%70%72%39%24%24%21%17%15%To manage and bookappointmentsTo allocate medicalresources in hospitalsTo

54、review scans anddiagnose illnessesTo make decisionsabout healthinsurance coverageFor deciding whatmedication toprescribeFor an initial primarycare consultation inplace of a doctorI wouldnt mind whether AI or a human didthis,as long as it was done wellI would prefer a human to do thisI would prefer A

55、I to do thisPreference of AI or a human to do each of the followingAI in healthcareAnd please say whether you would prefer artificial intelligence(AI)or a human to do each of the following in And please say whether you would prefer artificial intelligence(AI)or a human to do each of the following in

56、 healthcare:Base:All(n=5,074)For an initial primary care consultation in place of a doctorFor deciding what medication to prescribeTo make decisions about health insurance coverageTo review scans and diagnose illnessesTo allocate medical resources in hospitalsTo manage and book appointmentsI wouldnt

57、 mind whether AI or a human did this,as long as it was done wellAI in educationAI in education18AI is generally seen as acceptable for admin-based tasks in education,but not in replacement of teachersAI being used for tasks such as compiling research in academia(57%),creating lesson plans(56%),and m

58、arking work(54%)are considered acceptable by a majority of the public.Even over half(55%)of AI Abstainers AI Abstainers think it would be unacceptable to use AI for creating lesson plans,while 36%of this group think this is an acceptable use.It is considered less acceptable by the public for AI to b

59、e involved in teaching,with a majority of the public opposing it being used at each schooling level tested.That said,a third of the public would still find it acceptable for AI to be used to teach lessons in high school(32%)or university(34%).A majority of AI Optimists AI Optimists think it is accep

60、table for AI to be used in teaching at high school(54%)and university(53%).The public are split on using AI to review university applications,with 44%saying it is acceptable and the same proportion unacceptable.29%32%34%44%55%56%61%57%56%54%44%34%32%27%To compile research inacademiaTo create lesson

61、plansTo mark coursework/examsReviewing universityapplicationsTo teach lectures/seminarsin universityTo teach lessons in highschoolTo teach lessons inelementary schoolNet:AcceptableNet:UnacceptableAcceptable/Unacceptable for AI to doPlease say whether you think it would be acceptable or unacceptable

62、to use artificial intelligence(AI)in each of the followPlease say whether you think it would be acceptable or unacceptable to use artificial intelligence(AI)in each of the followinging aspects of aspects of education:Base:All(n=5,074)To teach lessons in high schoolTo teach lessons in elementary scho

63、olTo teach lectures/seminars in universityReviewing university applicationsTo compile research in academiaAI in education19Around half of the public either would prefer AI to create lesson plans(15%)or wouldnt mind whether AI or a human did this,as long as it was done well(34%),compared to 43%who wo

64、uld prefer a human to perform this task.Even AI AbstainersAI Abstainers are open to AI performing these tasks,with 29%not minding,as long as it was done well.The US public are far less likely to prefer AI over a human when it comes to teaching,with a clear majority preferring a human teaching at ele

65、mentary school(74%),high school(71%)or university(66%).A quarter(24%)of AI Optimists AI Optimists wouldnt mind AI teaching lectures and seminars at university if it was done well,with an additional 15%actually preferring AI to perform this task.One in five(20%)of the AI Ignorant AI Ignorant also wou

66、ldnt mind if AI or a human taught university lectures,as long as it was done well.The public are open to AIs use for admin tasks in education,but have a strong preference for humans to be used in teachingIn the qualitative groups,both AI Abstainers AI Abstainers and IgnorantIgnorant expressed concer

67、n that AI could lead to problems within the education system,especially when it comes to it enabling students cheating and harming the development of critical thinking skillsPreference of AI or a human to do each of the following17%16%15%14%8%7%6%39%43%43%52%66%71%74%34%33%34%26%18%15%12%To compile

68、research inacademiaTo mark coursework/examsTo create lesson plansReviewing universityapplicationsTo teach lectures/seminars inuniversityTo teach lessons in highschoolTo teach lessons in elementaryschoolI wouldnt mind whether AI or a human did this,as long as it was done wellI would prefer a human to

69、 do thisI would prefer AI to do thisI worry about students ability to write and form complex thoughts and ideas on their own without the support of the tech.-AI AbstainerMeanwhile,AI Optimists were cautiously hopeful about its potential benefits“AI,used well,can help tailor lessons to individual lea

70、rning styles.”AI OptimistAnd please say whether you would prefer artificial intelligence(AI)or a human to do each of the following in And please say whether you would prefer artificial intelligence(AI)or a human to do each of the following in education:Base:All(n=5,074)To teach lessons in high schoo

71、lTo teach lessons in elementary schoolTo teach lectures/seminars in universityReviewing university applicationsTo compile research in academiaI wouldnt mind whether AI or a human did this,as long as it was done wellAI in law and orderTalk to a researcherThe public think the use of AI in facial recog

72、nition and to review legal documents is acceptable,but not in place of a juryAI in law and order21The use of AI by police in facial recognition is seen as acceptable by 62%of the public.Amongst AI Abstainers AI Abstainers opinions are split,with 47%saying it is acceptable compared with 44%who say it

73、 is unacceptable.Half of the public also think it is acceptable to use AI to review and analyze evidence(49%)and to review legal documents(51%),while 45%think it is acceptable to use AI to decide where authorities should patrol.However,when it comes to replacing a jury in a trial,the overwhelming ma

74、jority say it is unacceptable(71%).Even AI OptimistsAI Optimists,who think it is acceptable to use AI across all other areas of law and order tested,think replacing a jury is unacceptable(61%).28%37%39%41%71%62%51%49%45%18%For use in facialrecognitionTo review legaldocumentsTo review and analyzeevid

75、enceTo decide where topatrolIn place of a juryNet:AcceptableNet:UnacceptableAcceptable/Unacceptable for AI to doIn the qualitative groups,all segments expressed significant concern about AIs usage in law enforcement.Many feared that AI would perpetuate the same biases they currently observe in polic

76、ing,especially if used to aid in the identifying of suspects.There are already biases in facial recognition tech,there are biases in the police departments,I could see all of that being amplified using AI-AI AbstainerIt will help automate some of the more tedious duties of officers.-AI OptimistAI AI

77、 Optimists Optimists saw some potential for AI to help law enforcement in administrative tasks,such as evaluating officer performance and speeding up paperwork.Please say whether you think it would be acceptable or unacceptable to use artificial intelligence(AI)in each of the followPlease say whethe

78、r you think it would be acceptable or unacceptable to use artificial intelligence(AI)in each of the followinging aspects of aspects of law and order:Base:All(n=5,074)In place of a juryTo decide where to patrolTo review and analyze evidenceTo review legal documentsFor use in facial recognitionB The p

79、ublic would overwhelmingly prefer a human jury,but are more open to its use in other aspects of law and orderAI in law and order2230%15%13%14%7%29%45%47%48%75%30%32%29%30%10%Use in facial recognitionReviewing legaldocumentsDeciding where policepatrolReviewing and analysingevidenceIn place of a jury

80、to decidethe outcome of a trialI wouldnt mind whether AI or a human did this,as long as it wasdone wellI would prefer a human to do thisI would prefer AI to do thisThree quarters of the public(75%)would prefer a human jury to decide the outcome of a trial over AI,while 7%would prefer AI and 10%would

81、nt mind either way.Across most other aspects of law and order asked about,the public are more open to AI being used.Three in ten wouldnt mind if AI or a human reviewed legal documents(32%),analyzed evidence(30%)or decided where police patrolled(29%)so long as it was done well.When it comes to facial

82、 recognition,the public are very much divided.Indeed,a similar proportion would prefer AI to do this(30%)as would prefer a human(29%),with the remaining(30%)not minding so long as it was done well.Even half ofAI Abstainers AI Abstainers would prefer AI to be used here(18%)or not mind if AI or humans

83、 performed this task(32%).Preference of AI or a human to do each of the followingAnd please say whether you would prefer artificial intelligence(AI)or a human to do each of the following in And please say whether you would prefer artificial intelligence(AI)or a human to do each of the following in l

84、aw and order:Base:All(n=5,074)In place of a jury to decide the outcome of a trialReviewing and analyzing evidenceDeciding where police patrolReviewing legal documentsI wouldnt mind whether AI or a human did this,as long as it was done wellAI in theworkplaceTalk to a researcherB 24AI in the workplace

85、16%62%8%Never acceptable to use AI in theworkplace,even if its more efficientSometimes acceptable to use AI inthe workplace,only if a human isntbeing replacedAlways acceptable to use AI in theworkplace,even if replacing ahumanSix in ten(62%)think that it is acceptable to use AI in the workplace,as l

86、ong as a human isnt being replaced,while 16%think it is never acceptable.Just 8%think it is always acceptable,even if replacing a human.AI OptimistsAI Optimists are most likely to think it is always acceptable to use AI in the workplace,even if it means replacing a human(17%),while virtually no AI A

87、I AbstainersAbstainers(1%)share this view.However,59%of AI AbstainersAI Abstainers still think it is sometimes acceptable to use AI in the workplace,as long as no humans are being replaced.When it comes to AIs impact on work/life balance,the public are split.Just under a quarter(23%)think it will im

88、prove it,with 34%saying it will make it worse.A further 14%say it will make no difference,while 29%are unsure.AI Optimists AI Optimists are most likely to think it will lead to a better work/life balance(45%),while AI Abstainers AI Abstainers are most likely to feel it will lead to a worse work life

89、 balance(52%).The AI Ignorant AI Ignorant are most likely to be uncertain(46%).Most of the public think it is at least sometimes acceptable to use AI in the workplace,though very few feel it always is23%34%14%29%BetterWorseNo impactDont knowAcceptable/Unacceptable for AI to be used in the workplaceI

90、mpact of AI on work/life balanceThinking about the use of artificial intelligence(AI)in the workplace,which of the following comes closest to your viewpoiThinking about the use of artificial intelligence(AI)in the workplace,which of the following comes closest to your viewpoint?nt?Base:All(n=5,074)W

91、hat impact,if any,do you think increasing artificial intelligence(AI)presence in the workplace will have on work/life baWhat impact,if any,do you think increasing artificial intelligence(AI)presence in the workplace will have on work/life balanlance for workers?ce for workers?Base:All(n=5,074)Always

92、 acceptable to use AI in the workplace,even if replacing a humanSometimes acceptable to use AI in the workplace,only if a human isnt being replacedNever acceptable to use AI in the workplace,even if its more efficientB Training people is considered the only acceptable use of AI in the workplaceAI in

93、 the workplace2538%54%54%56%51%34%34%31%To train peopleTo review jobapplicationsTo review requests forpromotions or pay risesReviewing misconductin the workplaceNet:AcceptableNet:UnacceptableJust over half of the public think it is acceptable to use AI to train staff(51%).The public are far less sup

94、portive of other workplace uses with a majority saying it is unacceptable for AI to be used to review job applications(54%),requests for promotions or pay rises(54%),and misconduct in the workplace(56%).Six in ten(60%)of AI Optimists AI Optimists consider using AI to review job applications compared

95、 to 26%who think it is unacceptable.Conversely,AI Abstainers AI Abstainers think this use is unacceptable by 66%to 27%.Acceptable/Unacceptable for AI to doIn the qualitative groups,AI Optimists and Ignorant suggested that AI could be used to make tedious,boring tasks faster and easier.They hope that

96、 it can be used to assist humans in being more efficient and effective at their jobs,and possibly enabling them to work less.If you think about it,technology developments have always allowed people to work less,like innovations in farming equipment,machines to create products.-AI IgnorantAI Abstaine

97、rs were more skeptical of AI in the workplace arguing that it can bring more frustration than it does benefits to the people its supposed to assist.Ive worked a variety of customer service jobs.In my experience,employees dont like the automated systems either.By the time the caller gets to you,theyr

98、e even more het up by the wait and navigating the endless menus.Humans have a need for authentic interactions.AI AbstainerPlease say whether you think it would be acceptable or unacceptable to use artificial intelligence(AI)in each of the followPlease say whether you think it would be acceptable or

99、unacceptable to use artificial intelligence(AI)in each of the followinging aspects of aspects of the workplace:Base:All(n=5,074)Reviewing misconduct in the workplaceTo review requests for promotions or pay risesTo review jobapplicationsB The majority would prefer a human to carry out all workplace t

100、asksAI in the workplace2610%10%13%11%64%63%51%51%17%19%28%31%Reviewingmisconduct in theworkplaceTo review requestsfor promotions orpay risesTo review jobapplicationsTo train peopleI would prefer AI to do thisI would prefer a human to do thisI wouldnt mind whether AI or a human did this,as long as it

101、 was done wellJust under two thirds(64%)would prefer a human to review misconduct in the workplace compared to 10%who would prefer AI to perform this task and 17%who wouldnt mind,as long as it was done well.A similar proportion(63%)would prefer a human to review requests for promotions or pay rises(

102、63%),with half saying the same for reviewing job applications(51%)and training people(51%).Around three in ten say that they wouldnt mind if a human or AI trained people(31%),as long as the job was done well.This increases to 38%for AI Optimists AI Optimists with an additional 21%of this group favor

103、ing AI over a human for this task.The majority of AI Abstainers AI Abstainers would prefer humans to perform all tasks tested,particularly when reviewing promotions(82%)or misconduct(83%).Preference of AI or a human to do each of the followingAnd please say whether you would prefer artificial intell

104、igence(AI)or a human to do each of the following in And please say whether you would prefer artificial intelligence(AI)or a human to do each of the following in the workplace:Base:All(n=5,074)B B A majority of the public think AI in the workplace will result in fewer jobs,but those in work think it

105、is unlikely their job will be replacedAI in the workplace27In the qualitative groups,there was agreement across segments that AI will result in fewer jobs in the future.However,AI Optimists AI Optimists and AI Ignorant AI Ignorant envisioned a future where AI creates new jobs as it makes others obso

106、lete,arguing that it could create new opportunities as it takes others away.10%67%12%12%Net:More jobsNet:More jobsNet:Fewer jobsNet:Fewer jobsNo differenceNo differenceDont knowDont know29%Of those in work think it is likelylikely that AI will replace their job59%Of those in work think it is unlikel

107、yunlikely that AI will replace their jobWhether AI will result in more/fewer jobsLikelihood that AI will replace jobs(of those in work)I think about the industrial revolution and how people lost factory jobs,but I wouldnt necessarily go back and change that because that came with a lot of benefits.I

108、 think jobs will change,but it might not be all bad.AI IgnorantEvery technology has“displaced”workers.That is a given.The key is to retrain them while supporting them.AI OptimistTwo thirds(67%)think that AI will result in fewer jobs,while 10%feel it will lead to more jobs,and 12%think it will make n

109、o difference.Even 57%of AI OptimistsAI Optimists think it could result in fewer jobs;this rises to 89%of AI AbstainersAI Abstainers.Despite this,the majority(59%)of those currently in work think it is unlikely AI will replace their job,and just 29%think it is likely they will be replaced by AI.In yo

110、ur view,do you think the use of artificial intelligence(AI)in the workplace would result in more or fewer jobs,or woIn your view,do you think the use of artificial intelligence(AI)in the workplace would result in more or fewer jobs,or woulduld it make no difference?it make no difference?Base:All(n=5

111、,074)How likely or unlikely do you think it is that artificial intelligence(AI)will replace you in your job?Base:All who work(n=2,554)Regulationof AITalk to a researcherA majority of the public think that regulations are needed for AI to continue to be developed Regulation of AI29AI Abstainers AI Ab

112、stainers are more likely to think that AI should either not be allowed to continue to be developed with existing uses permitted(12%)or banned altogether(23%).However,half of AI Abstainers AI Abstainers(50%)still think that AI should be allowed to be continued to be developed,with regulations in plac

113、e.Similarly,half(51%)of the AI Ignorant AI Ignorant believe that AI should continue to be developed with regulations in place,while AI Optimists AI Optimists are most likely to think AI should be allowed to continue to be developed either without any regulations(11%)or with regulations in place(78%)

114、.5%AI should be allowed to continue to be developed without any regulations60%AI should be allowed to continue to be developed with regulations in place8%AI should not be allowed to continue to be developed,but existing uses should be permitted9%AI should not be allowed to continue to be developed,a

115、nd all existing uses should be banned4%said None of these,and 13%Dont know Free developmentBanned use of AIThinking about any future regulation for artificial intelligence(AI),which of the following,if any,comes closest to your Thinking about any future regulation for artificial intelligence(AI),whi

116、ch of the following,if any,comes closest to your vieviewpoint?wpoint?Base:All(n=5,074)B B Half of the US public think AI companies should be allowed to give their opinion,but not have a deciding role in constructing AI regulationRegulation of AI30Whether AI companies should be involved in regulation

117、In the qualitative groups all segments expressed a large amount of skepticism over AI companies intentions with their product.They did not trust that most AI developers had the general publics best interests at heart,and many did not want them involved in regulating AI.“I want to believe AI develope

118、rs want to help people but thats never the case It always comes down to money,the more money they can make for other the better for them.AI AbstainerWhether because of generalized distrust in the government or because of doubts of its ability to effectively regulate AI,most were not confident in the

119、 governments ability to regulate AI.AI Ignorant and AI Optimists expressed a desire for an independent body made up of experts to establish AI regulations.I would like more governmental influence but Im not sure they even understand the tech.AI Ignorant18%16%51%15%AI companies should be involved in

120、all steps to construct theregulationAI companies should be allowed to give their opinions,but shouldnot be given decision making responsibility in constructing theregulationAI companies should not be involved at all in constructing theregulationDont knowA further 15%think AI companies should be invo

121、lved in all steps to construct the regulation,something a quarter of AI Optimists AI Optimists agree with(25%).A similar proportion(16%)think that AI companies should not be involved in the construction of regulation at all,a feeling more strongly felt by AI AbstainersAI Abstainers(31%).Nonetheless,

122、just under half of AI AbstainersAI Abstainers(47%)think that AI companies should be allowed to express their opinion.Thinking about any future regulation of artificial intelligence(AI),to what extent do you think that artificial intelligencThinking about any future regulation of artificial intellige

123、nce(AI),to what extent do you think that artificial intelligence(e(AI)companies should or should not be involved in the construction of this regulation?AI)companies should or should not be involved in the construction of this regulation?Base:All(n=5,074)MethodologyTalk to a researcherMethodologyYouG

124、ov conducted an online survey with 5,074 US adults from November 10-29,2023.The sample was structured and weighted to be representative of the population by age,gender,region,education and race.The survey covered a range of topics,including:Perceptions of the most important issues facing the USGener

125、al perceptions of AIPerceptions of the use of AI across key sectors,including:Health,Education,Law and order,The workplace.Regulation of AIThe first-stage qualitative online communities were used to inform the content and structure of the quantitative survey.Following the survey,YouGov ran statistic

126、al segmentation analysis,revealing three segments in the population.Quantitative Qualitative To understand how the public thinks about AI,YouGovs qualitative experts hosted an online community with 30 participants from September 11-15,2023.The data from this community informed the structure and cont

127、ent of the quantitative survey.Following the segmentation analysis,YouGovs qualitative experts conducted 3 online focus groups(one with each segment)from February 5-6,2024.In the groups,we spoke to 33 people from a range of different demographic backgrounds,ensuring a spread by gender,age,region,soc

128、ial grade and ethnicity.The discussion followed the topics in the survey to understand general perceptions,benefits and drawbacks,and regulation(among other topics).All quotes in the report come from the qualitative focus groups.Get in touch*Statistical segmentation analysis analyzed a data set of o

129、ver 5,000 individuals with complete responses to all segmentation survey items.The underlying factor variables were computed using polychoric and continuous correlations,oblimin rotations,the minimum residual estimation,and the ten Berge method for generating scores.The clustering procedure used the

130、 k-means algorithm and Euclidean distances,paired with the k-means+initialization algorithm.The selection of the solution rested on a variety of compactness measures,such as silhouette width and within-cluster variance,alongside subjective appraisals in terms of interpretability and strategic useful

131、ness-all results have been weighted.32Get in touchCustom Research34YouGov Custom Research is a full end-to-end service with our experts,from sample framing and questionnaire design to analysis and presentations.Get deep dive research and customized insights to inform your strategy.Find the most accu

132、rate and efficient way to reach the data you need,fast,with flexible,rich,and unique solutions.Receive custom reports tailored to your needs or access advanced analytics in our data visualization platform.Bespoke market research and intelligence27 million+registered panel members in 59+marketsGlobal

133、 reachTailored industry and sub-sector specific adviceBespokeSpecialists in research methodologyExpertiseExplore Custom ResearchYouGov,2024,all rights reserved.All materials contained herein are protected by copyright laws.Any storage,reproduction or distribution of such materials,in whole or in par

134、t,in any form without the prior written permission of YouGov is prohibited.This information(including any enclosures and attachments)is propriety and confidential and has been prepared for the exclusive use and benefit of the addressee(s)and solely for the purpose for which it is provided.We make no

135、 representations,warranties or guarantees,whether express or implied,that the information is accurate,complete or up to date.We exclude all implied conditions,warranties,representations or other terms that may apply and we will not be liable to you for any loss or damage,whether in contract,tort(including negligence),breach of statutory duty,or otherwise,even if foreseeable,arising under or in connection with use of or reliance on the information.We do not exclude or limit in any way our liability to you where it would be unlawful to do so.Thank youLiving consumer intelligence|