Understanding Older Adults Perceptions and Challenges in Using AI-enabled Everyday Technologies

2025-05-06 0 0 276.6KB 12 页 10玖币

侵权投诉

arXiv:2210.01369v1 [cs.HC] 4 Oct 2022

Understanding Older Adults’ Perceptions and Challenges in

Using AI-enabled Everyday Technologies

Esha Shandilya

School of Information

Rochester Institute of Technology

Rochester, NY, USA

es4524@rit.edu

Mingming Fan∗

Computational Media and Arts Thrust

The Hong Kong University of Science and Technology

(Guangzhou)

Guangzhou, China

Division of Integrative Systems and Design

The Hong Kong University of Science and Technology

Hong Kong SAR, China

mingmingfan@ust.hk

ABSTRACT

Artiﬁcial intelligence (AI)-enabled everyday technologies could help

address age-related challenges like physical impairments and cog-

nitive decline. While recent research studied older adults’ experi-

ences with speciﬁc AI-enabled products (e.g., conversational agents

and assistive robots), it remains unknown how older adults per-

ceive and experience current AI-enabled everyday technologies in

general, which could impact their adoption of future AI-enabled

products. We conducted a survey study (N=41) and semi-structured

interviews (N=15) with older adults to understand their experi-

ences and perceptions of AI. We found that older adults were en-

thusiastic about learning and using AI-enabled products, but they

lacked learning avenues. Additionally, they worried when AI-enabled

products outwitted their expectations, intruded on their privacy, or

impacted their decision-making skills. Therefore, they held mixed

views towards AI-enabled products such as AI, an aid, or an adver-

sary. We conclude with design recommendations that make older

adults feel inclusive, secure, and in control of their interactions

with AI-enabled products.

CCS CONCEPTS

•Human-centered computing →Empirical studies in acces-

sibility.

KEYWORDS

AI-enabled everyday technologies, Older Adults, Interview, Per-

ceptions

ACM Reference Format:

Esha Shandilya and Mingming Fan. 2022. Understanding Older Adults’ Per-

ceptions and Challenges in Using AI-enabled Everyday Technologies. In

∗Corresponding Author

Permission to make digital or hard copies of all or part of this work for personal or

classroom use is granted without fee provided that copies are not made or distributed

for proﬁt or commercial advantage and that copies bear this notice and the full cita-

tion on the ﬁrst page. Copyrights for components of this work owned by others than

the author(s) must be honored. Abstracting with credit is permitted. To copy other-

wise, or republish, to post on servers or to redistribute to lists, requires prior speciﬁc

permission and/or a fee. Request permissions from permissions@acm.org.

Chinese CHI 2022, October 22–23, 2022, Guangzhou, China and Online, China

ACM ISBN 978-1-4503-9869-5/22/10. . . $15.00

https://doi.org/10.1145/3565698.3565774

The Tenth International Symposium of Chinese CHI (Chinese CHI 2022), Oc-

tober 22–23, 2022, Guangzhou, China and Online, China. ACM, New York,

NY, USA, 12 pages. https://doi.org/10.1145/3565698.3565774

1 INTRODUCTION

Artiﬁcial Intelligence (AI)-enabled products, such as online shop-

ping and autonomous driving, are becoming increasingly integrated

into daily lives that people are often unaware of their presence

and potential impact on daily lives [45, 66]. Recent research began

to understand people’s experiences and attitudes toward AI and

showed that age may aﬀect their attitudes and experiences with

AI [9, 45, 66]. While younger adults tend to encounter AI products

more often and are generally positive about AI, older adults aged

60 years or above [20, 42, 46] tend to have relatively less experience

with AI products [9] and may face challenges adapting and using

AI products [1]. As AI could potentially help older adults deal with

age-related issues such as physical impairments [25], cognitive de-

cline [8], and emotional isolation [67], it is crucial to understand

older adults’ personal encounters and experiences with AI to better

inform the design of AI technologies for this population.

Recent research began to explore older adults’ experiences with

speciﬁc AI-enabled products, such as conversational agents (e.g.,

Alexa, Google Home) [2, 3, 27, 35, 44, 49, 58], smart appliances (e.g.,

smart vacuum cleaner) [30, 43, 48, 51, 56], and assistive robots [26,

27, 50, 63]. While experiences with speciﬁc AI-enabled products

are imperative to improve these products, the general perceptions

and understanding of AI-enabled products aﬀect older adults’ ac-

ceptance and expectations of current and future AI-enabled tech-

nologies in general. Furthermore, it remains largely unknown i)

how older adults perceive the term “AI" and expect AI-enabled

technologies to work, ii) their knowledge of AI and trust in AI,

and iii) the reasons why they might have speciﬁc (mis)perceptions

about AI.

In this research, we aim to understand qualitatively older adults’

knowledge, experiences and (mis)perceptions of using AI-enabled

products by exploring the following research questions (RQs):

•RQ1: What are older adults’ personal experiences with AI-

enabled products?

•RQ2: Do older adults face any challenges in using AI-enabled

products? If yes, what are they?

Chinese CHI 2022, October 22–23, 2022, Guangzhou, China and Online, China Esha Shandilya and Mingming Fan

•RQ3: What are older adults’ perceptions of the AI technol-

ogy in AI-enabled products?

We ﬁrst conducted an online survey study with 41 participants

to screen and recruit older adults (60 or above) who have expe-

rience in using AI-enabled products. Through our survey study,

we found that despite older adults’ limited knowledge of AI, they

used a variety of AI-enabled products, though some products like

customer-service chatbots, recommendation features, and voice-

assistants were used less frequently than other ones like search

engines and navigational maps that use AI in the background. To

gain a deeper understanding of older adults’ responses to the sur-

vey questions, we recruited 15 survey respondents who had some

experience of using AI-enabled products for interviews. Our in-

terview ﬁndings show that older adults’ limited awareness of AI

might be attributed to the lack of a tech-savvy community and the

negative portrayal of AI in news and media that discouraged them

from exploring AI-enabled products. Moreover, other adversarial

behaviors or consequences of using AI-enabled products such as

AI stealing their data, creating smarter scams, and owning their

data further made them hesitant to explore AI-enabled products.

Lastly, we discuss design considerations based on older adults’ ex-

periences and challenges in using AI-enabled products. In sum, we

make the following contributions in this work:

•An initial understanding of older adults’ personal experi-

ences and challenges in using AI-enabled products;

•A qualitative understanding of older adults’ general percep-

tions and understanding of the AI used in AI-enabled prod-

ucts;

•Design considerations to make AI-enabled products more

acceptable and accessible to older adults.

2 BACKGROUND AND RELATED WORK

2.1 Older Adults and Technology Use

Technological products facilitate us from ordering food, shopping,

managing health care to ﬁnancial and social security services [12],

and weaves into the lives of all, including older adults, aged 60

or above [20, 42, 46]. Keeping pace with the changing technology

becomes crucial for oneself [31]. Adapting to the rapidly changing

technology is often left to an individual’s technological capabilities

and skill-sets [1]. Melenhorst et al. [38] argue that speciﬁcally older

adults deter to adopt new advancements in technology not because

of usability issues or cost, but because of their lack of awareness

about the beneﬁts or value the technology could provide. Older

adults’ motivation to use and invest their energy in new technol-

ogy depends on their knowledge of its perceived advantages and

worth [12, 38].

Digital literacy is of paramount importance for an individual

to determine the perceived use and value of a technology [1]. Ac-

cording to Leahy and Dolan [31], digital literacy means individu-

als’ ability to discern the usefulness of a particular technology and

their comfort in using it for various purposes; it applies to any-

one from a non-technological background, including older adults.

However, to attain this skill and proﬁciency, technology should

be usable and accessible to everyone [10]. Older adults experience

anxiety, fear, lack of control, and unawareness of existent services

while using any technological product [1]. Additionally, they also

encounter ergonomic[29] and usability issues [4, 14, 36] in techno-

logical products like online banking [22],mobile phones [28, 29],

tablet computers [60], and websites [4, 14, 36].

To summarize, past work suggest that an older adult’s motiva-

tion to use and adapt an advancing technology depends on three

factors: i) technology’s perceived usefulness and potential [12, 38],

ii) digital literacy to experience technology’s beneﬁts [1], and iii)

personal apprehensions in using digital technologies [1]. However,

these factors were mostly derived based on non-AI technologies.

As recent advancements have been continuing introducing AI into

various technological products, we are yet to understand how and

in what forms older adults’ motivations manifest into (un)adoption

and usage of AI.

2.2 Older Adults and AI

AI technology has become part and parcel of our daily lives [33].

AI has revolutionized many sectors like the e-commerce indus-

try [52, 57], from providing personalized recommendations while

shopping [52], virtual shopping assistants to help us with queries

[57], to detecting frauds [41]. Additionally, AI is extensively used

in healthcare sectors for early disease diagnosis [54], in social me-

dia to enhance users’ engagement [21]. Many times, users are not

aware that they interact with products that use AI, which aﬀects

their perception and may create false expectations from AI sys-

tems [34, 45]. Therefore, Long and Magerko [34] emphasize the

importance of AI literacy for users and that digital literacy is the

prerequisite for it.

AI has been used in enhancing the lives of older adults [2, 3,

18, 26, 48]. Older adults experience various age-related issues such

as loneliness, functional decline impacting their physical and men-

tal abilities hindering them from performing their daily life activ-

ities comfortably [16]. Consequently, most of the research stud-

ies aim at resolving/assisting older adults with age-related issues

through speciﬁc AI-enabled technological interventions [6, 13, 16].

For example, how conversational agents like Alexa, smart appli-

ances like a smart cleaner [51], or assistive robots augment their

physical or cognitive abilities [8, 24, 25, 40, 43, 63], alleviate the

social isolation felt by older adults [11, 35], and their feelings to-

wards it [3, 44, 49, 56, 58]. However, we still do not know older

adults’ general perceptions, experiences, and concerns related to

AI-enabled technologies, which could ultimately play a signiﬁcant

role in setting their expectations of AI products and their decisions

to embrace them.

A few recent studies [9, 45, 53, 66] studied people’s awareness,

applications, challenges, beneﬁts, and drawbacks of AI. Holder et

al. [9] conductd a survey study with more than 2000 UK respon-

dents and found that more than 74% of the older adults aged 55

years and above have limited AI knowledge and limited exposure

to AI products as compared to younger adults [9]. However, none

of these earlier studies [9, 45, 66] were oriented towards study-

ing older adults’ perceptions of general AI technology, experiences

and needs in using an AI-enabled product(s) eﬀectively. Further,

these studies have not focused on how older adults’ perceptions

impact their usage or vice-versa. Therefore, we take a step further

to understand older adult’s perceptions of AI, experiences in using

AI-enabled products, and their expectations from such products

Older Adults’ Perceptions and Challenges with AI-enabled Everyday Technologies Chinese CHI 2022, October 22–23, 2022, Guangzhou, China and Online, China

through qualitative research methods by focusing on general AI

technology.

3 ONLINE SURVEY STUDY

To understand older adults’ experiences in using AI-enabled prod-

ucts and their perceptions about AI technology, we ﬁrst conducted

an online survey to recruit participants aged 60 or above, a com-

mon age threshold for older adults used by United Nation, World

Health Organization and prior research [20, 42, 46]. In this section,

we describe our survey methodology and survey ﬁndings that in-

formed the follow-up interview study. We further discuss the in-

terview study in Section 4 and interview ﬁndings in Section 5. All

studies were approved by IRB at our institution.

3.1 Materials and Procedure

To identify older adults aged 60 or above with experience in us-

ing AI-enabled products, we designed 7 survey questions (6 close-

ended and 1 optional open-ended). We did not provide a formal

deﬁnition of AI in the survey for two reasons. First, we would like

to keep it open so as to elicit older adults’ understanding of AI.

Second, a formal deﬁnition of AI might be too technical to be un-

derstood by all potential older adult participants. Instead, we used

11 AI-enabled services as examples to probe their perceptions and

understanding of AI: i) customer service chatbots, ii) email spam

ﬁlter, iii) navigational maps like Google Maps, iv) search engines

like Google, v) video recommendation features on YouTube, Net-

ﬂix, vi) online shopping recommendation features on Amazon, vii)

personal ﬁnance like mobile-check deposit, viii) ride-sharing appli-

cations like Uber, ix) robotic vacuum e.g. iRobot, x) social-network

applications like Facebook, and xi) voice-assistants like Alexa. Our

questions were centered around collecting: i) demographic infor-

mation (older adults’ age and gender), ii) frequency of AI-enabled

products used by older adults, iii) their perceived AI knowledge

level, and iv) their sources, if any, used to learn AI. The survey

also allowed participants to sign up for a follow-up interview by

providing their email-ids.

We advertised our survey on social advertisement platforms (e.g.,

Craigslist, Facebook, Reddit) to target older adults from the North-

American region. We also contacted local senior communities and

adopted the snowball sampling technique during a six-week period

to collect responses. The advertisements directly deﬁned our tar-

get participant group—older adults— and the purpose of the study.

Participants completed the questionnaire in their own time and

entered a raﬄe to win a $30 award.

3.1.1 Survey Participants. We received 41 survey responses in to-

tal but removed 6 because they did not acknowledge being an older

adult (60 years or above). All the survey responses gathered were

from North-American countries. Among the remaining 35, 27 re-

ported their demographic information: gender (16 females, 10 males,

and 1 preferred not to say) and age (6 between 60 and 69, 13 be-

tween 70 and 79, and 8 were 80 or above). A total of 23 respon-

dents signed up to participate in a follow-up interview, and we

eventually recruited 15 of them who reported using at least two

AI-enabled products for the interviews.

Next, we brieﬂy report our survey ﬁndings to provide an overview

on: i) self-reported knowledge level of AI, ii) AI learning resources

used, and iii) frequency of use of AI-enabled products by the sur-

vey participants.

3.2 Survey Results

We asked the respondents to report their perceived AI knowledge

level by choosing a score on a scale of 1 to 5 (where 1 = ‘I am

not aware of anything about AI.’ and 5 = ‘I am very knowledgeable

about AI.’), as shown in Table 1. Most of our survey participants,

20 out of 35 participants, reported having some knowledge about

AI. We then presented a list of AI learning sources as shown in Ta-

ble 2 with an “other” option to the respondents to select or provide

sources that they used to learn anything about AI. The most used

learning source is “Internet search” followed by “Friends/Family

members” and “Newspaper”; apart from the listed source, older

adults also reported other learning sources like radio and news

channels.

Our survey ﬁndings showed that participants used a variety of

AI-enabled products. The most frequently used AI-enabled prod-

ucts are search engines like Google, 27 out of 35 participants use it,

followed by navigational applications like Google Maps (25 out of

35) and email spam ﬁlter (16 out of 35) and online shopping recom-

mendation feature (16 out of 35). However, respondents occasion-

ally used customer service chatbots (7 out of 35), recommendation

systems on video platforms such as Netﬂix (11 out of 35) and on-

line shopping sites like Amazon (10 out of 35), voice-assistants like

Alexa (11 out of 35), and none of our participants (0 out of 35) used

Robotic Vacuums.

Table 1: This survey data shows the self-reported AI knowl-

edge levels (i.e., scores and descriptions) among older adults.

Score Perceived AI Knowledge Level No. of re-

sponses

1 I am not aware of anything about AI. 0 (0.00%)

2I am aware of AI but do not have much

knowledge about it.

12 (34.28%)

3I have some knowledge about AI. 20 (57.14%)

4I am knowledgeable about AI. 3 (8.5%)

5I am very knowledgeable about AI. 0 (0.00%)

3.3 Summary of Survey Findings

Our survey revealed that most of our survey respondents prefer

using various AI products irrespective of their age and have some

knowledge of AI. Moreover, respondents leverage search engines,

newspapers, and depend on families and friends to learn about AI.

However, we do not yet know the reasons for older adults’ lim-

ited knowledge of AI technology and their sparse use of certain AI

products.

The survey study helped us in identifying the older adults who

use AI-enabled products but, we still did not have a deep under-

standing of i) the experiences of older adults in using AI-enabled

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

arXiv:2210.01369v1[cs.HC]4Oct2022UnderstandingOlderAdults’PerceptionsandChallengesinUsingAI-enabledEverydayTechnologiesEshaShandilyaSchoolofInformationRochesterInstituteofTechnologyRochester,NY,USAes4524@rit.eduMingmingFan∗ComputationalMediaandArtsThrustTheHongKongUniversityofScienceandTechnology(Gu...

展开>> 收起<<

Understanding Older Adults Perceptions and Challenges in Using AI-enabled Everyday Technologies.pdf

共12页,预览3页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Understanding Older Adults Perceptions and Challenges in Using AI-enabled Everyday Technologies

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: