what canadians think about artificial intelligence and … · 2018. 8. 30. · what canadians think...

What Canadians Think About Artificial Intelligence and Implications for Canadian Businesses December 1, 2017

E X P E C TAT I O N S A N D I M P L I C AT I O N S F O R G O V E R N M E N T

I M P L I C AT I O N S F O R B U S I N E S S E S

I M P L I C AT I O N S F O R B R A N D S A N D M A R K E T E R S

D O : A D O P T I O N A N D U S A G E

F E E L : E M O T I O N S A N D C O N C E R N S

T H I N K : E X P E C TAT I O N S A N D C O M P R E H E N S I O N

Contents

1 9

2 2

2 3

1 1

8

3

What Canadians Think About Artificial Intelligence and Implications for Canadian Businesses 2

Executive Summary Artificial intelligence has reached a turning point. Every day, there are new

advancements in AI, new ways for companies to use AI, new assessments of

threats and opportunities of AI. As more and more businesses gear up to adopt AI

technologies, do they truly understand how consumers perceive AI and what they

expect from it? How can brands apply these insights to solve their tough business

challenges? Given consumer perceptions, how can marketers influence innovation

and drive ethical standards around AI technology?

To understand how Canadians perceive AI, including what they think, feel, and do, we

conducted a survey in August 2017 with 1,001 people representing the general adult (18+)

population of Canada.

As you will soon learn, Canadians anticipate the advent of the AI economy with

wonderment, but also with concern. This paper provides important insights and lessons

into how government, businesses, and marketers can work to alleviate the trust issues and

build a bridge to the future where AI is a positive, trusted, and valuable component of

everyday products and services.

T H I N K F E E L D O

3

Canadians have high expectations of AI Like much of the world, the Canadian public has not escaped the AI fervor. In fact, AI has

become part of our social discourse with 52% of Canadians saying they have heard news or

participated in conversations about artificial intelligence in the past year. Though gender

differences suggest that men are more attuned to this AI news than women (63% for men

vs. 42% for women), there are no age differences as all generations of adult Canadians

reported similar exposure to the topic.

Given this level of awareness, it should not come as a surprise that expectations of AI are

also high. Half of Canadians (49%) think they use or will use AI-enhanced devices and tools

in their personal life in the next 5 years. Expectations are even higher among Millennials

(currently aged 18 to 34), of whom 58% already use or plan to use AI-enhanced devices in

their personal life in the next 5 years.

A significant percentage of Canadians (61%) think “AI will completely change the way we

live and work”. On the other hand, only 23% think “AI has nothing to do with the way I live

and work and is not likely to affect me in the future”. Except for a small minority,

Canadians are clearly listening for and anticipating the effects that AI will have upon their

daily lives.

What Canadians Think About Artificial Intelligence and Implications for Canadian Businesses

W H AT C A N A D I A N S T H I N K O F A I

4

AI understanding is surprisingly strong Do the media over-hype expectations for AI? Do Canadians genuinely understand

what artificial intelligence is, or what it can and cannot do?

When asked, 45% of Canadians reported that they understand current levels of AI

developments “very well or somewhat well”. More specifically, men reported higher

comprehension levels than women (55% vs. 35%), as did Millennials (aged 18 to 34)

in comparison to Baby Boomers/Silent Generation (aged 50 to 64) (57% vs. 37%)

(see Figure 1).


45

55

35

57

48

37

21

Total Male Female 18-34 35-49 50-64 65 or older

Figure 1. Understanding of current level of AI development (very/somewhat well) % Q: How well do you understand the current level of Artificial Intelligence development?


5

Canadians understand and anticipate AI capabilities It is, of course, one thing to say you understand, and a completely different thing to prove

you understand. To test this theory, we offered participants a series of potential AI

capabilities to decipher their true comprehension of AI, and Canadians passed the test.

The majority of participants were able to identify the most common uses of AI, e.g.,

recognize speech (72%) or translate languages (68%). And, they rejected the (current)

unrealistic applications, e.g., make moral choices (7%) or read people’s minds (5%) (see

Figure 2).

A small minority (15%) think that AI has already surpassed human intelligence, while

half (51%), similar to the experts in the field, think that AI will catch up to humans in 10

to 50 years.

Finally, when asked what artificial intelligence is, a majority (55%) correctly identified AI as

“self-learning algorithms designed to do a specific task”. It is alarming, however, that 42%

incorrectly indicated AI is “a self-aware computer program.”



6



While this understanding of AI remains mainly non-technical (only 14% had even heard the

terms "machine learning" or "deep learning"), it does point to high levels of awareness and

anticipation among Canadians.

72

68

58

56

44

44

37

33

24

24

21

16

15

11

9

7

5

14

Figure 2. Understanding of current level of AI abilities, % | Q: To the best of your knowledge, at its current state of development, which of the following things can Artificial Intelligence do?

Recognize speech

Translate between languages

Recognize pictures of animals and human faces

Recommend songs you’ll like

Spot burglars in your home

Detect a range of medical conditions from images

Trade stocks

Make a medical diagnosis

Interpret and understand the world around it

Replace you in your job

Handle insurance claims

Understand your personal problem and give you some advice

Laugh at a joke

Write sports articles for the Associated Press

Decide on court cases

Make moral choices

Read people’s minds

I don’t know/None of these

7 What Canadians Think About Artificial Intelligence and Implications for Canadian Businesses

Women and men comprehend AI equally As we saw earlier, men reported much higher levels of understanding Artificial Intelligence

than did women (55% of men vs 35% of women said they understand it very/somewhat

well). Digging deeper into the topic, the data showed that while women report lower levels

of understanding, they actually have the same level of comprehension - a very real

example of the Think versus Do premise.

When offered a series of potential AI applications, women correctly identified the

capabilities of current AI technologies with the same degree of accuracy as did men - with

one exception. Men were more likely than women to say that AI can trade stocks (43% vs.

31%).

In terms of defining AI, women were as likely as men to think that AI is “self-learning

algorithms designed to do a specific task” (54% of women vs. 56% of men), but less likely

to think that AI is a “self-aware computer program” (38% of women vs. 46% of men).

Whether men over-estimate their level of understanding or women under-estimate theirs,

the answer to that question is yet to be resolved.


8

H O W C A N A D I A N S F E E L A B O U T A I


Canadians feel both wonderment and concern about AI Given relatively (and unexpectedly) high levels of comprehension of AI technology, it

should come as no surprise that Canadians do not fear AI.

From a set of positive, neutral, and negative emotions, participants were asked to select all of

the emotions that reflected their immediate reactions towards the words “Artificial

intelligence.” Feelings of curiosity were selected by 71% of participants (see Figure 3). Positive

feelings were chosen by about half of participants (47%), including 24% who chose anticipation

and 21% who chose excitement. On the other hand, about 40% of participants chose negative

feelings, including 34% who chose concern and 18% who chose fear/anxiety.

46

32

15 15

Mys

tery

/unk

now

n

Scie

nce

fictio

n

Surp

rise

Conf

usio

n

Figure 3. Immediate reaction to Artificial Intelligence, % | Q: And more specifically, which of the following words and phrases describe your immediate reaction when you hear the words 'Artificial Intelligence'?

71% NET Curious

24 21 19

8 7

Antic

ipat

ion

Exci

tem

ent

Awe

Conf

iden

ce Joy

47% NET Positive

34

18 11

3 2

Conc

ern

Fear

/Anx

iety

Hype

Disa

ppoi

ntm

ent

Bore

dom

40% NET Negative



Though fear and anxiety were included in this motley collection of feelings, glaring

negative emotions (e.g., fear) were in the minority. The most common reactions - mystery,

science fiction, anticipation, and excitement - coalesce to create an overriding sense of

wonderment with a side helping of concern.

With those feelings of wonderment and the lack of fear, it is interesting that 77% of

Canadians are ready to use a personal AI assistant. In fact, Canadians are keen for an

assistant to have human characteristics such as a sense of humour (71%), a name (68%),

high ethical standards (68%), and values and worldviews that match their own (62%). In

essence, they are keen for an AI assistant that can engage with them on a human,

emotional level.

Job security and privacy are the top two worries When asked about potential negative outcomes of AI over their lifetime, a large majority of

participants worried about decreasing levels of privacy (75%) and increasing levels of job loss

(72%) (see Figure 4). These two negative effects generated the largest gender differences

among all of the options. Specifically, women were more likely than men, by as much as nine

percentage points, to single out these worries. There were, however, no age differences.


Societal concerns do not translate to personal concerns Delving deeper into these two concerns, job security presents a paradox. A large majority

of people worry that AI will put people out of work (72%) but, at the same time, only 14%

believed that they themselves might lose their job due to AI in the next five years (plus 3%

who said that they had already lost their job to AI). Canadians know that AI will have major

effects on employment, but they might not be ready to acknowledge its impact on their

own employment.


75

72

62

58

52

48

29

18

Figure 4. Potential negative effects of AI | Q: Thinking about all potential negative effects Artificial Intelligence can have, how likely are the following scenarios to happen during your lifetime?

Increasing lack of privacy

Putting a lot of people out of jobs

Destroying a lot of businesses (e.g., retail)

Total control/surveillance of the citizens by the government

Creating more inequality

Changing global balance of power

The rise of the robots and enslavement of humanity

Eliminating the human race

11

W H AT C A N A D I A N S D O W I T H A I


This feeling of personal job security may be partly due to the fact that 49% of Canadians

believe AI technologies will lead to “increased employment through a new ‘knowledge’

industry” and 54% expect AI to result in “increasing productivity and overall country’s

wealth”. In other words, Canadians know that AI will have major impacts on employment

but they feel that, in the end, those impacts will be positive.

When it comes to personal security in relation to AI, about a third of people admitted to

having concerns. Specifically, about 37% of participants feel annoyed that AI technology

“collects too much information" about them. The same proportion (37%) worries that “AI

in my phone (e.g., Siri, Google Assistant) constantly listens to my conversations” and 22%

feel concerned “about machines/artificial intelligence uncovering my deepest secrets”.

Canadians may have resolved concerns about their own employment prospects but

they have yet to resolve concerns about their own privacy.

Canadians are ready to use AI technologies in their personal life Canadians reported a rather high level of comfort with using AI technologies for various

managerial aspects of their lives. They’re quite content to use AI when it comes to

scheduling and bookings (70%), recommendations for shopping (59%), and home utilities

management (73%).


However, comfort levels decrease in areas that directly affect their personal lives and well-

being (see Figure 5). In fact, the tasks that generated the lowest comfort levels were also

the tasks that generated the largest differences of opinion between men and women.

Participants were least comfortable with using AI to diagnosis medical issues without

human involvement (50% for men vs. 37% for women), drive vehicles (45% for men vs.

33% for women), and interact with children (42% for men vs. 35% for women). Essentially,

people are happy to use AI to deal with things, but when it comes to the personal health

and safety of “me and my loved ones”, people aren’t yet convinced.

73

70

59

59

59

56

53

52

49

49

43

39

38

Figure 5. Comfort level with AI in various areas of life, (somewhat/very comfortable) % Q: How comfortable would you feel if an AI would…

Regulate the temperature in your home (based on your habits)

Organize and notify you about your schedule and appointments

Control your house utilities and appliances

Provide you with recommendations on what to purchase

Provide you with recommendations on what to eat

Provide you with financial advice

Arrange your travel plans

Answer your calls to customer service

Send personal messages to people in your contact lists

Provide you with legal advice

Diagnose your medical condition (without doctor’s involvement)

Drive you or your family (without human involvement)

Interact with and teach your children



Canadians are prepared to upend the traditional Customer Decision Journey

Earlier, we learned that about 77% of Canadians would use an AI assistant for shopping.

This includes 17% of people who are prepared to completely hand over the entire

shopping experience to AI. It also includes 28% of people who would use AI to find the best

price, and 32% who would use AI to discover options from which they can choose. Given

that major portions of marketing and advertising processes are dependent on these

components, the Customer Decision Journey is headed for change – fast!

At the same time, it’s interesting that 23% of people can’t imagine using AI for shopping

when many commercial websites already use AI, or precursors of AI, to recommend

products and services as part of the product shopping experience or as part of their

marketing and adverting engines.

Despite high expectations and comfort, adoption of AI technologies is still in its infancy

We know that Canadians are aware of the types of AI technologies that currently exist and

they’re prepared to use them. However, only 12% of participants reported that they

currently use AI technologies in their personal life right now. Similarly, only 11% said they

use them at work right now.

That isn’t to say Canadians aren’t open to acquiring AI technologies. Offered the choice

between buying an AI-powered product and a standard product at the same price, more

than one third (37%) of participants opted for the AI-enhanced one. Clearly, Canadians are

ready to embrace AI technologies. They just haven’t - yet.



Lack of clear unique benefit and trust are barriers to AI adoption Besides using AI technologies without realizing it (e.g., product predictions, text or song

prediction, smart home devices), low adoption rates may be the result of not knowing the

specific benefits that AI-powered products already offer. We see this in the fact that the AI

technologies Canadians already use are the technologies also viewed as most useful. For

example, “promptly getting answers to my questions” (46%) is an ability that people

already have on their computers and phones (see Figure 6).


46

43

43

35

34

32

28

26

24

22

21

18

18

13

Figure 6. Most useful personal applications of AI, % | Q: Which abilities of Artificial Intelligence would be most useful to you personally?

Promptly getting answers to your questions

Health assessment/diagnosis

Managing home utilities to ensure savings

Self-driving cars

Helping make best financial choices

Managing everyday tasks

Simplifying shopping

Increasing your productivity

Creating a personalized music list for you

Getting prompt help from an AI CS rep

Arranging your travel plans

Developing tailored learning program

Helping make best shopping decisions

None


Low adoption rates are also likely the result of a lack of trust. While many companies are

excited to take advantage of the benefits of new AI-based marketing technologies, many

consumers currently feel unprotected from and exploited by companies trying upsell or

seemingly mislead them with those technologies.


37

37

37

30

25

25

24

1

9

Figure 7. What consumers find annoying with online purchase recommendations, % | Q: What are the most annoying things about companies providing you with recommendations online based on your purchase history? Please select no more than 3.

My purchase or browsing history is being used without my approval

They show me the products they want me to buy, not what I want to buy

They collect too much information about me

The recommendations follow me anywhere I go online

They don’t have a good understanding of my needs, show me the products I don’t need

It is not clear if the recommended products have best price or over-priced

They show me the products I already purchased

Other

None of these



Consumers want AI technology that works for them, not just for companies. Recommendation engines are an obvious example of how technologies have caused

consumers to be disappointed in AI technologies and mistrust companies. Canadians are

generally fine with AI algorithms offering shopping recommendations (52%), but many

people (37%) find it annoying that those same algorithms “show me the products they

[the company] want me to buy, not what I want to buy”. Even now, shoppers are

regularly annoyed when they continue to be shown products they have already

purchased (see Figure 7).

Сhatbots are another example of AI technologies that have demonstrated lackluster

success thus far. Chatbots are quickly taking over customer service roles, and most

consumers are okay with that (52%). However, consumers demand to know (78%), and

rightly so, whether they are talking to a human being or a bot.

With so much uncertainty around AI technologies and the potential to abuse them, many

people are suspicious of companies that implement AI, and negative associations (73%)

of those companies outweigh positive ones (58%) (see Figures 8 and 9). Indeed, 41% of

people believe that companies using AI are “focused on reducing their costs at the

expense of people”, while only 12% agree that “they are focused on providing the best

product and service to customers”. Consumers want AI technology that works for them,

not just for companies.



29

22

16

12

12

12

11

11

Figure 8. Positive reactions towards companies using AI, % | Q: Many companies are moving towards using Artificial Intelligence in their operations and communications with customers What do you think about them?

They are the future

They are not afraid to invest in new technologies

They are/will be making my life easier

They are focused on providing the best product and service to customers

They are cool and innovative

They create more high-tech / high paid jobs

They contribute to the advancement of our economy overall

They are focused on providing the best value to their customers

58% NET Positive



41

36

29

18

13

11

3

Figure 9. Negative reactions towards companies using AI, % | Q: Many companies are moving towards using Artificial Intelligence in their operations and communications with customers What do you think about them?

They are focused on reducing their costs at the expense of people

They steal jobs from people

I am worried that they might misuse a powerful AI to their own advantage

They don’t have any ethical standards for Artificial Intelligence in place

They contribute to the demise of human race

They create unhealthy competition

Other (please specify)

73% NET Negative

19

E X P E C TAT I O N S F O R G O V E R N M E N T


The government is expected to play an important role AI technologies are unquestionably driving social change, and Canadians are looking to the

government to be a part of and help lead this change.

The majority of Canadians expect the government to play a role, particularly in the

regulatory area. Specifically, 58% of Canadians want the government to “develop safety

guidelines for using AI technologies”, and 54% of Canadians expect the government to

“create new laws and regulations with regards to using AI technologies” (see Figure 10). On

the other hand, only 11% of Canadians do not want the government involved in any way in

the process of unleashing AI technologies on society.

Deve

lop

safe

ty g

uide

lines

fo

r usin

g AI

tech

nolo

gies

Crea

te n

ew la

ws a

nd re

gula

tions

with

re

gard

s to

usin

g AI

tech

nolo

gies

Prov

ide

free

trai

ning

pro

gram

s for

pe

ople

pus

hed

out o

f wor

kfor

ce b

y

AI te

chno

logi

es a

nd a

utom

atio

n

Crea

te m

ore

educ

atio

nal a

nd tr

aini

ng

prog

ram

s to

ensu

re w

e ha

ve e

noug

h

AI sp

ecia

lists

Heav

ily re

gula

te th

e AI

fiel

d

Plac

e he

avie

r tax

es o

n th

e co

mpa

nies

th

at su

bstit

ute

hum

an la

bour

w

ith A

I and

aut

omat

ion

Inve

st in

the

AI se

ctor

to m

ake

sure

our

co

untr

y ha

s a c

ompe

titiv

e ad

vant

age

Prov

ide

loca

l AI c

ompa

nies

w

ith ta

x br

eaks

The

gove

rnm

ent s

houl

d no

t be

invo

lved

in a

ny w

ay

58 54 48 39 35 28 28 7 11

Figure 10. What the government should do with AI, % | Q: How comfortable would you feel if an AI would…


While few people are aware of specifics around the Canadian government’s current

investment in AI (9%), the majority (84%) support this investment. However, most

support this investment with at least some concerns (58%). If these investments are to

serve the public well and continue to generate broad support, the government must

clearly communicate the benefits to the public.

Given that most public concerns about artificial intelligence are centered on job

security, the government must come up with clear programs that both create “more

educational and training programs to ensure we have enough AI specialists” (38%

agree) and provide “free training programs for people pushed out of workforce by AI

technologies and automation” (48% agree).

Indications are that the government endorses AI technologies, but if they decide to

also use AI systems, they must ensure those systems retain a human face when

communicating with the public. We have learned that Canadians are generally

comfortable with AI customer service representatives helping them to book tickets or

hotels, but only 14% would be comfortable if an AI agent answered their government

service inquiries. This concern comes back to a lack of trust in AI technologies that

affect people’s personal health and safety.

E X P E C TAT I O N S F O R G O V E R N M E N T


Canadians are rooting for their country to be at the front of AI research and development. The government must facilitate our potentially painful transition to the knowledge economy.

Lessons for the government:

1. Invest in AI Development: Invest in AI that reflects the public’s

desire for a brighter future in the knowledge economy.

2. Educate and Create Jobs: Create job transition plans and safety nets

for workers displaced by AI technologies.

3. Create Regulatory Boundaries: Create new laws and regulations

with regards to AI development and use, particularly with respect to

the privacy and security of personal information.

I M P L I C AT I O N S F O R G O V E R N M E N T


With high levels of awareness, comprehension, expectations, and readiness for adoption, businesses must create socially responsible AI technologies.

Lessons for businesses:

1. Develop AI Purposefully: Build algorithms that best serve

customer needs and make meaningful differences in both jobs

and personal lives.

2. Use AI Smartly: Minimize pain points and build trust. Consider

broader applications, and work to create industry standards rather

than company-specific standards.

3. Adopt AI Responsibly: Broaden the embodiment of social

responsibility to encompass more than just supporting a good

cause or using green technologies. Social responsibility must now

also include understanding how business strategies affect society

and developing strategic plans that maximize the positive effects

while minimizing the negative impacts of AI.

I M P L I C AT I O N S F O R B U S I N E S S E S


The way consumers engage with brands is going to change rapidly which will greatly impact how we market. Brands and marketers must strive for purposeful and ethical AI technologies that promote public trust.

Lessons for the brands and marketers:

1. Foster Consumer-Driven Innovation: Shape the end-user

experience in a positive way by satisfying specific needs rather

than simply using AI for 'upselling.’

2. Establish High Ethical Standards: Personalize content such that it

is useful and impactful for consumers not simply a manipulation

of human psychology. Rather than using AI technologies to

maximize short-term gains leading to increased customer

attrition, create AI technologies that help to build trust and foster

long-term customer satisfaction and engagement.

3. Be Transparent and Honest in Communications: Be prepared to

raise business standards and increase transparency around how AI

technologies are used. Take advantage of AI capabilities to

dramatically improve customer service and engagement.

I M P L I C AT I O N S F O R B R A N D S A N D M A R K E T E R S

Point of contact for inquiries:

Marina Laven | Director T 416.488.6422 x 240 [email protected]

2200 Yonge Street, Suite 400 Toronto, Ontario Canada M4S 2C6 www.sklarwilton.com

How is your business addressing

consumer perceptions of AI? Are you

prepared to face the changes and

challenges AI presents?

Sklar Wilton & Associates will guide

your business through the disruptive

force of AI and help solve your tough

business challenges.

what canadians think about artificial intelligence and … · 2018. 8. 30. · what canadians think...

Documents