Advances in Social Science, Education and Humanities Research, volume 631

Proceedings of the 2021 International Conference on Social Development and Media Communication (SDMC 2021)

The Effect of Artificial Intelligence on

the Human Idea of Free Will
Eric Luo*
The Webb Schools in Claremont, California, USA
*
Corresponding author. Email: [email protected]

ABSTRACT
Artificial Intelligence (AI) has made advances in its ability to mimic and surpass human problem solving and
decision-making abilities. As AI technology advances, the public begins to use AI products for their own personal
decision-making. However, AI products may encounter the problem of humans’ fondness of free will. Human free
will is traditionally conceived of as a kind of power to control one’s choices and actions. This research attempts to
find out how humans react to AI decision-making products, discover the underlying reasons relating to human free
will, and discuss future trends. Not only does this research help to give guidance to the development of AI products
that would be accepted within different human groups, but it also helps to understand how decision-making AI may
affect human culture and human society. In order to glean this, we have split this study into four parts. These are about
what decisions AI already make for humans, what they could possibly make in the future, how different people view
decision-making AI, and finally how humans may react to AI in general. This research starts from an analysis of the
paradox of choice, which guides that, in certain circumstances, less choice may lead to more satisfaction during the
decision-making process. In addition to the research about AI decision-making and the paradox of choice, surveys of
17 people were conducted as well as interviews of four individuals from different age and cultural groups. The
analysis showed a strong correlation between the age and cultural backgrounds and the responses to the acceptance of
AI decision-making. The conclusion of the investigation shows that in the future a more pronounced paradox of
choice would occur. While people stating that they value free will, breaking it down may show that people don’t
really need free will to be happy if the result is satisfactory. There was large disagreement as to whether AI decision-
making is positive or negative in human society. Future studies may do interviews and surveys at a larger scale to
glean more significant information and more research can be done into the ethical considerations of how humans
value free will.

Keywords: Artificial Intelligence, Human free Will, Paradox of Choice, AI Decision Making, Choice
overload, AI’s Effect on Humans’ free will, AI’s influence on human lives, Future trends in AI, User
Profiling, AI Apps, AI Products.

In the future and arguably in the present, there may

1. INTRODUCTION
be additional advances in AI technology that will allow
In recent times, Artificial Intelligence (AI) has made it to break into the mainstream and allow more people to
great advances in its ability to mimic human problem- utilize the decision-making abilities of AI for personal
solving and decision-making skills. Certain examples use. This AI could use data available in an increasingly
currently include applications in law firms, detecting connected world to make decisions based on enormous
financial fraud and making business decisions. AI has amounts of training data gleaned from the internet. At
been increasingly helping humans make decisions and, this point, larger tasks such as determining good career
in some cases, make decisions for humans. It has also paths and helping with socialization may be done by AI.
been shown that AI is superior to humans in certain If AI starts controlling people’s lives, playing the
decision-making, like those relevant to the stock market. game of life better than humans can, a question of free
will emerges. Many of the decisions which humans
make will now be made by something else, which may

Copyright © 2022 The Authors. Published by Atlantis Press SARL.

This is an open access article distributed under the CC BY-NC 4.0 license -http://creativecommons.org/licenses/by-nc/4.0/. 1355
 Advances in Social Science, Education and Humanities Research, volume 631

lead to people feeling losing freedom. In general, 2. THE PARADOX OF CHOICE

humans inherently like free will. If AI takes it away,
there may be problems relating to humans feeling In terms of determining how humans may react to
anxiety and depression. AI making decisions for them, it is important to see how
humans feel about choice and freedom.
To see this issue in action, chess can be analyzed. In
1997, IBM’s Deep Blue computer defeated reigning The term “Paradox of Choice” was coined by Barry
champion Gary Kasparov, in a highly publicized match. Schwartz in his 2004 book, which covered this topic [2].
After that, more powerful chess engines were This book combines studies from different areas of
developed, including Stockfish, which is currently one psychology. Schwartz describes the way humans often
of the most powerful chess engines. However, even feel more stress and anxiety when given more choice.
though humans have been rendered totally obsolete in When describing “why we suffer”, he says that choices
chess, humans still enjoy chess and play it. In fact, the that were once made by family, the community, and the
engines have become used more as a tool for making workplace, are now being done on an individual basis
humans enjoy chess more. Chess engines have been which puts extra pressure on people. He also writes that
banned as their use is considered cheating in many chess missed opportunities cause stress since every decision
circles [1]. Chess was and is still relegated to human will have trade-offs. If one makes decision A, they may
affairs where humans play as a form of leisure. People be unhappy with the trade-offs involved even though
did not see a large amount of anxiety and depression decision B may have other trade-offs which they
from the AI being better than them and replacing their haven’t considered.
decision-making. However, this is because chess is a In 2000, a landmark paper released by Sheena
game where people play others to challenge themselves Iyengar and Mark Lepper showed the paradox of choice
and have fun rather than solely winning. As Mr. in an experimental form [3]. Two shops were set up in
Wagonfuhr said, “If the AI knows exactly how your the “Jam Experiment”, both selling jams of different
traps work and how to prevent them, it makes the game flavors. One shop had only 6 choices while the other
uninteresting in comparison to the drama and shop had 30 choices. Out of the two shops, the shop
excitement of watching the opponents’ eyes look at the with more choices had 50% more people enter it than
pieces, hoping he didn’t see your trap.” the shop with fewer choices. However, only 3% of those
People’s lives are different, and “winning” is the end who entered the shop with more choices made a
result for most people. However, it should be kept in purchase compared to 30% in the shop, which had fewer
mind that humans could start seeing life as a game, not choices. Another significant statistic deduced from this
for survival and luxury but for fun. This is because an study is the average number of jams flavors sampled by
AI dominated world could optimize the supply chain the customers of both shops: 1.4 for the shop with less
allowing people to live longer and more luxuriously in choice and 1.5 for the shop with more. This shows that
spite of any decisions that they make. oftentimes, the human psyche prefers less choice even
though they think they will be happier with more
The question is whether humans will surrender choice. Hence the shop with more choice had more
control of their own lives to AI or reject AI completely. customers. The hypothesis discussed in the study was
If humans refuse to submit to AI, AI may need to that the immensely large number of choices paralyzes
include choice in its design, illusory or real, to make people to pick from and decide not to choose at all. This
appropriate decisions. That question can be further is supported by the fact that the number of samples
subdivided into four additional questions. taken by each customer was similar, suggesting that
1. What human decisions are already being made or those in the shop with more choices didn’t have enough
influenced by AI? information to decide which one they liked the most.
2. What AI could make human decisions in the There have been many attempts to duplicate the
future? paradox of choice in numerous studies. This has had
mixed success with large variances that suggest that
3. What is the opinion of people from different age choice overload and the paradox of choice could be tied
groups or cultural backgrounds on using AI to make to certain pre-conditions.
decisions concerning smaller and larger issues such as
career paths and relationships? A 2015 meta-analysis by Alexander Chernev
attempted to identify those pre-conditions [4]. It was
4. Based on the current psychological and shown that the complexity of the decision, the
sociological trends, how may humans react to AI taking difference in choice, the amount of uncertainty or
away free will? knowledge the decision maker has when processing the
benefits and drawbacks of each choice, and the extent to
which the decision maker aims to minimize the
cognitive effort required are reliable proportional

1356
 Advances in Social Science, Education and Humanities Research, volume 631

predictors of choice overload and the paradox of choice. adult with a Chinese background. These interviews were
The situations in which people minimize cognitive done in a semi-structured format. The interview began
effort was described in the meta-analysis as including by asking the participant to what extent they valued
decision intent (buying or browsing), decision focus their free will and why. After that, they were introduced
(choosing a group or an individual option), and the level to the topic of free will and encouraged to converse
of construal. In other words, the more complicated it is freely with the interviewer in relation to this. To follow,
to make the choices, the more likely it is that people will the interviewer presented the participant with different
feel overwhelmed. In this meta-analysis, choice scenarios and asked them their opinions. Each scenario
overload was measured by the amount of satisfaction involved a decision, and they progressively got more
and regret reported, the amount of choice deferral, and significant. The scenarios included the use of AI in
the amount of switching. picking restaurants, the use of AI for playing chess
competitively, the use of AI in choosing a career, and
In conclusion, the paradox of choice shows that
the use of AI in social and relationship decisions. The
humans are often paralyzed and stressed by large
variety of scenarios given reflects the different ways in
amounts of choice, especially when it is complicated to
which AI decision-making may affect humans. The
consider every choice’s drawback and benefit. This
interviewees are encouraged to ask questions
creates a more pressurized decision-making process.
throughout.
3. HYPOTHESIS On top of these interviews, 17 individuals took part
in a survey. The questions were closed versions of the
Based on the knowledge about the paradox of interview questions. The surveys were used to find
choice, it is hypothesized in this paper that humans will trends and correlations, while the interviews were used
not be opposed to AI making decisions for them and to justify them.
that AI may lead to a higher standard of living and stress
reduction. Aside from that, a case study was done on the
paradox of choice, which was done using jelly beans
In terms of how different age and cultural bought at a local supermarket. These jelly beans were
backgrounds may affect people’s opinions, it is used as an incentive for the surveys and interviews.
hypothesized that younger individuals will be more There were five different flavors of jelly beans in the
opposed to the idea of their free will being taken packet. A third of the test subjects were allowed to
because, in general, younger people strive for freedom choose between all five flavors, another third were only
from society and people of authority such as teachers or given two choices, and another third were given no
parents. In terms of cultural opinions, it is hypothesized choice on the flavor whatsoever. The test subjects
that there will be a negligible effect. However, there weren’t told what each jelly bean’s flavor was and were
may be a slight correlation between a western allowed to ask what the flavors were. The test subjects
background and an opposition to AI taking free will. were then surveyed on how satisfied they were with
their choice. They were also asked whether they would
4. METHODOLOGY or wouldn’t want to choose the same flavor again.

To determine what decisions AI has already started

making for humans, library research and fieldwork were
conducted in the form of interviews and observations
relating to the everyday decisions of an individual.
Notes were written every time a decision was made in
reference to or prompted by an AI or computer
algorithm.
To assess what decisions AI may be able to make in
the future, library research examined the opinions of
current futurologists and the technologies that already
exist, some well-established and others in early
development. In addition, interviewees were asked to Figure 1. Jelly Bean Experiment
share their opinions on future possibilities. To consider how humans in general will react to AI
Four interviews were conducted to ascertain how in the future, a general analysis was done on all of the
age and culture affect people’s opinions on the use of AI information gleaned from the previous methods of
in career planning, relationships, and smaller daily research.
decisions. The interviewees were a teenager with a
western background, a teenager with a Chinese
background, an adult with a western background, and an

1357
 Advances in Social Science, Education and Humanities Research, volume 631

5. ANALYSIS because of the immense psychological triggers

employed in these apps.
5.1. What decisions have AI already started Another example is the use of AI to help choose a
making for humans? restaurant or food venue. Apps such as Yelp or
Dianping in China take into account the user’s previous
Currently, AI usually uses machine learning to
preferences, the frequency of visits or orders at a
perform tasks. They use training data to adjust their
restaurant, the location of the user, and the reviews the
parameters to function appropriately for their assigned
user gives to determine the restaurants they recommend.
task. This kind of AI already does many things better
These kinds of apps often use advertisements paid for
than humans. Companies have used AI to collect more
by the restaurants that want to encourage new
personal data than what is humanly possible, sometimes
customers. While the human still has a choice, certain
more than an individual’s family and friends know
psychological cues such as the order, the number of
about them. For example, in 2012, popular retailer
stars, or other distinguishing marks can all drastically
Target sent advertisements and coupons for baby
increase or decrease the chances of an individual
products in the mail to a teenage girl [5]. Her father ran
choosing a certain restaurant.
into the local Target in a rage, asking whether Target
was trying to get his daughter pregnant. However,
Target wasn’t trying to encourage her to become
pregnant. She was already pregnant. Target had used AI
algorithms to tailor their marketing campaigns to
relevant consumers. These algorithms used data gained
from the teenage girl’s shopping choices to determine
that she was pregnant. They found that she was buying
certain products such as unscented lotion and soaps and
cotton balls, which were correlated with pregnancy.
Target’s algorithms connected this with pregnancy and
suggested the baby products as ideal items to sell to the
girl in question.
There are already a huge number of decisions made
by AI right now. This study mostly studies the
implications of AI in individual lives rather than for
business, so the examples examined in this study will be
predominantly based on individuals.
One of the most publicized ways that AI has affected
humans is in recommending media for people. As a Figure 2. Examples of Psychological Triggers on the
famous example, the YouTube “algorithm” decides Popular Chinese Food Review Service, Dianping
what videos are recommended on the front page of
AI has also helped people determine which college
YouTube for different people. It looks at what people
or college major to study. There is a myriad of online
have clicked on before and what they are most engaged
tools to help with that choice. As an example, The
by to determine what kinds of videos they highlight to a
Council of University Presidents of Thailand has
user. This has been proven to be extremely effective.
developed an AI system to analyze user interests and
Another example is the Chinese app, Toutiao, which
match them with university majors [6]. What is
recommends personalized articles to people based on
interesting about the goals of this AI is that the
their interests. These applications are often very
developer, Mr. Suchatvee, said, “For many Thai kids,
addictive and can lead to a reliance on AI to stimulate
parents play a major role in their decision making about
pleasure in the brain. One adult subject said during the
careers and study. Students need guidance while they
interview, “Toutiao makes it hard to steer away from
navigate a difficult series of choices. The kindest thing
what it recommends for you. The more you use it, the
is to support them rather than taking those choices
narrower and homogenous it becomes. When I first
away.” He believes that giving children support when
started using it, I hated how it didn’t give me a chance
making such important decisions is better than having
to choose what I wanted to watch, but eventually, I
parents decide.
surrendered.” Because of human psychology and
passiveness, people will often click on whatever is Numerous decisions AI currently makes instead of
recommended instead of searching the wide variety of humans. These other decisions include the selection of
choices available in a place such as a library. Even if an dating partners, the picking of houses to rent or buy, and
individual wants to, they may give in to passiveness the picking of movies. In the business world, there are
also decisions such as firing and hiring for jobs. Many

1358
 Advances in Social Science, Education and Humanities Research, volume 631

start-ups are working on AI, which can replace middle- likely possible for AI to work without human support in
management positions. It usually uses AI to determine the future.
which tasks can be done by an AI and which tasks
If these two milestones, theory of mind and
require internet freelancers. The AI decides who to hire
individual functionality, are achieved, then many more
and who to fire. While the workers work, the AI learns
tasks could be done by AI. AI could make decisions on
from them, slowly learning how to do the task. In fact,
socializing, gaining new relationships and friendships. If
one study predicted that AI will do 69% of managers’
AI managed to achieve a theory of mind, this is
work by 2024 [7]. On the other side of the college
guaranteed to happen. In this scenario, it is possible that
admissions system, many college admission offices are
AI could determine how good of a fit someone would be
employing AI to filter out unqualified applicants. In
in a dating app scenario and exactly how the
general, many decisions are already being advised on or
relationship or friendship would go. Following the AI
influenced by AI, though there aren’t many decisions
would almost guarantee a healthy and stable relationship
done entirely by AI yet.
with someone else since the AI can look at vast amounts
of data that is put online. There will be even more data
5.2. What decisions could AI make in the for the AI to learn from in the future, which will lead to
future? smarter AIs. AI could also help people converse more
productively. If the AI wants an individual to get to
In the future, AI will make further advancements
know somebody, then the AI can also help with the
that will enable them to make more decisions on behalf
actual conversation and interaction. For example, when
of humans, perhaps even replacing the need for human
speaking, the AI could determine how someone feels
support altogether. AI doesn’t need to be perfect. They
about them and what types of words would work the
just have to be close to or better than humans.
best. With a theory of mind, the AI could easily use
For AI to make further progress, certain milestones immense processing power to do this. There is much
need to be achieved. The examples to be discussed here evidence that a theory of mind could be achieved in the
involve the AI’s ability to have the theory of mind and future [9, 10].
organise data without humans’ supervision.
AI could also help people in college applications [6].
The theory of mind is defined as imagining other College consulting is a multi-billion-dollar industry in
people’s thoughts, understanding that they know and the United States. As discussed previously, there are
feel different things. Human children usually develop already certain programs that help people decide which
this at around the age of four. A test called the Sally- majors or colleges to apply to. The future may bring
Anne test determines whether children have developed additional advancements that could act as college
this skill [8]. Basically, there is one box and one basket. consultants, analyzing huge amounts of publicly
Sally puts a cookie in the box. Sally leaves, and Anne available data to determine which kind of students
puts the cookie in the basket. When Sally returns, the would get into which colleges. It can also determine
subjects are asked where Sally will look. According to what kind of extracurricular activities are suitable or
people with a theory of mind, Sally will look in the box what essays would help. Because AI has unsupervised
because she doesn’t know that the cookie switched learning abilities as well as a theory of mind, it could do
places. However, people without a theory of mind will this. However, it would be much more efficient if the
say that Sally will look in the basket because they don’t future brings a completely AI-based college admission
understand that Sally doesn’t know exactly what they, office as it has already started to do. AI is more
the observer, do. To make decisions, theory-of-mind AI predictable than human beings making the theory of
that can understand humans must be created. They will mind ability of the AI more pronounced. Because such
also be able to recognize emotions. AI has been shown an AI would make the admissions process unfair, the
to have the basic theory of mind, meaning that they can admission office may try different methods to make sure
recognize the way other AI and human brains work. its admission office is fair and more unpredictable.
Eventually, an arms race between the AIs would
The way many AI and machine learning algorithms
develop, creating a Nash equilibrium where neither side
work right now is based on a supervised model. For
can improve their chances. The ball has fallen into a
example, if an AI must recognize hand-written digits, it
stable state. This is a possibility of such a system.
is given a series of pictures of handwritten digits, which
are classified. However, things aren’t always classified AI could help determine what kinds of jobs or life
in the real world, and humans must make sense of raw paths an individual should take. If AI has higher
data. This is currently one of the separations between AI computing power and access to data, it may find
and human brains. AI must be able to take raw input correlations in data that can determine which kind of
data and make sense out of it to achieve its goals. There people will prefer what paths in life and in careers.
is currently much development occurring in the These types of AI can be similar to the college
unsupervised learning AI. This shows that it is most consulting AIs in which they determine ideal majors or

1359
 Advances in Social Science, Education and Humanities Research, volume 631

jobs. These AIs could look at a person’s personality and expansive. They may believe that they want to try out
other details gathered from the internet to find out what many different things.
types of jobs would be more suitable. If the computing
The reason why people are all opposed to the use of
power allows for rudimentary simulations, it can even
AI in decision making in social life is that people often
tell an individual what exactly may happen in their lives
find the fun in socializing.
if they choose a certain path. This type of AI would be
similar to what human parents do with children in The result of the cultural differences may be caused
certain cultures. by the fact that Chinese people put less value on
freedom in their culture while westerners value
In general, AI could make decisions for normal
freedom. For example, the American government has
people in many ways if certain technological goals are
put stricter restrictions on face recognition software and
achieved. These examples can be further subdivided
technology because it believes it infringes on privacy.
into smaller examples. In general, those technological
While there is some opposition to its use in civilian
goals can be achieved based on current rates of
circles, it is mostly accepted in China.
development.
The amount of support for AI’s use in social life was
5.3. What is the Opinion of Different People similar to the amount of support for AI in general,
from Different Age Groups or Cultural suggesting a connection between the two. It could be
Backgrounds on the Use of AI to Make possible that people associate their overall decisions
Decisions on Both Small Daily Decisions and with social decisions. This requires further
investigation.
Larger Decisions on Career Paths and
Relationships with Other People?
5.4. Given Current Trends, How May Humans
Based on the survey, it was found that teenagers and React to AI Taking Away Free Will?
adults are similar in their thoughts regarding free will
and the use of AI to replace a free will. However, when These studies were referenced above in the “Paradox
zooming in on smaller decisions about restaurants, of Choice” section and were all done to help marketing
adults were more likely to use it. For the chess initiatives. However, the conclusion can help determine
competition question, there were almost no affirmative the amount that humans may oppose the use of AI to
answers except for a single teenager. This can be make decisions for them. In the future, there will be
considered an outlier. In terms of career planning and more decisions humans need to make. This is also due
college majors, teenagers were very opposed to using AI to the access to more choices in their lives as social
for this. Adults were more open to the idea. For the mobility increases. There may also be more additional
socialization and relationships question, they were both choices from the cultural changes that are happening. As
similarly opposed to the idea. Based on this, it can be discussed in Barry Schwartz’s “Paradox of Choice”,
concluded that Teenagers and Adults have similar levels decisions that used to be made by the family or the
of support for most forms of AI aside from the use in community or the workplace are now being made by the
career paths and relationships. individual. The amount of choice overload is further
compounded by the fact that humans may have less time
Based on the survey, it was found that individuals to consider each decision and more information due to
with a Chinese background valued freedom and free the internet and big data. As discussed in Chernev’s
will less than western individuals. However, they had meta-analysis, the ability of the decision maker to make
similar levels of support for the use of AI in general, the decisions and the complexity of choice are strong
which were low. Those of Chinese background were determiners of the amount of choice overload. In
more supportive of the use of AI in making small addition, because more things that an individual does
decisions. Both Chinese and Western people were in will be published online, decisions will have more value
opposition to the use of AI for winning chess and more importance as everyday decisions carry more
competitions. Chinese people were more welcome to the weight.
idea of using AI to help with career planning and life
paths. They were all similar for the use of AI in helping Based on the trends shown by the survey, it may
with socializing and relationships. mean that there is a lot of disagreement on the use of
this technology and that humans may need larger
It may be hypothesized that teenagers were opposed amounts of cultural change for it to be possible for AI to
to the idea because adults may investigate the past as make decisions for humans.
something which can’t change and wish that they made
better decisions. This may be why Ms. Tian said that Based on this, it may be determined that psychology
people shouldn’t take so many tangents and that they shows us that the paradox of choice may have an effect
should spend their time on one thing and do it well. and that humans may be able to adapt to having less
Teenagers see the future as something more fluid and choice in their lives. However, the jelly bean experiment

1360
 Advances in Social Science, Education and Humanities Research, volume 631

shows that a small amount of choice will lead to much decisions for him. However, after further digging, he
more satisfaction. This means that, in general, it may be was a lot more indecisive and, in the end, he said, “Ehh,
preferred to have a smaller amount of choice and that AI who cares, I’ll do what I feel like doing”. This subject
may need to program in the illusion of choice. mentioned both the Thunderhead in the book Scythe and
the book I Have No Mouth and I Must Scream. This
6. CONCLUSION subject isn’t quite sure about whether he thinks AI
taking free will help humanity or imprison humanity.
The advent of AI will have large ramifications on
people’s lives, and as it becomes more advanced, it may Subject 2: Mr. Sheng
begin guiding humans in their lives as advisors. Age: 15
However, humans may dislike this system as many Cultural Background: Chinese
humans believe that free will is important. This paper
has discussed whether such ideas are possible and how In general, this subject wasn’t opposed to the idea of
humans may react to such a development. It has been AI taking away free will. However, for scenarios
concluded that AI will most likely be able to accomplish involving making friends and relationships, he was
such a task as playing the game of life. It has been opposed. He was also opposed to the use of AI for chess
concluded that people may adapt to this new because he thought that long term decisions should be
environment well, but a small amount of real or illusory done by AI, though he also thought that restaurants
choices will make people happier. This research may could be done by AI.
have ramifications on any AI which is used in an Subject 3: Ms. Tian
advisory way as the choice may need to be programmed Age: 31
into many programs to make customers happier. Cultural Background: Chinese
However, the lack of freedom shouldn’t be a large
problem. In the future, more research can be done into This subject had a lot of thoughts on the topic and
the ways AI may interact with each other or how spoke for over two hours. In general, this subject was
humans may integrate their minds with AI. In addition, more practical, mostly caring about AI taking away her
more research can be done into the ethical “physical and financial security”. In terms of playing
considerations of how humans value free will. While chess, she was opposed to doing it and believed that AI
people may say that they value free will, breaking it can’t do things for you, it can only cover you up. She
down may show that people don’t need free will to be often referenced “some people” in a negative light. She
happy if the result is satisfactory. believes in practicality and that going on tangents
because of personal interest is bad. She trusts
ADDENDUM technology more than other people. She is a STEM
major as well as an investor in technology which may
Raw Data and Interview Notes be the reason for this. She said, “I trust AI. I trust data. I
trust logic.”
1. Interviews
Subject 4: Mr. Luo
Four people were interviewed, and the summaries of Age: 25
the interviews are as follows. Cultural Background: Western, American
Subject 1: Mr. Wagonfuhr
Age: 14 This subject was more opposed to AI taking away
Cultural Background: Western, German-American free will. He said that people feel better when they have
free will and that freedom is a fundamental way of
In general, this subject was a lot more thoughtful living. However, when zooming in on the use of AI for
about the topic, asking a lot of clarifying questions. This restaurants, he often referred to using AI to narrow
subject appears to deny the fact that AI would be able to down choices instead of using it to have only one
achieve this, often asking questions such as “How does choice. In general, he believed that keeping choice or
it know that?”. He is more utilitarian and wasn’t the illusion of choice was incredibly important.
opposed to the use of AI and is fine with others making

1361
 Advances in Social Science, Education and Humanities Research, volume 631

2. Survey
Table 1. The following information has the answers grouped by age and ethnicity. The numbers will represent the
percentage of people who answered “Yes”. There were eight adults and nine teenagers who took the survey. Of those
eight adults and nine teenagers who took the survey, ten were Chinese and seven were Westerners.
Questions Chinese Western Teenager Adults Total
Do you value free will and freedom? 60 86 78 75 70
If an AI could make decisions about your life for you and tell you what would happen 50 57 56 50 53
if you made a certain decision, would you use it?
Let’s suppose that there was an AI that could tell you which restaurant you should go 90 71 78 88 82
to. Current apps only tell you if you want to eat Chinese food, the best Chinese food
places. However, this AI can tell you exactly what you want before you know
yourself. It would know based on all the data you give it from the internet and big
data. Would you use this?
Suppose you are at a chess competition. There is a technology which can tell you 10 0 11 0 5
through the glasses that you are wearing, which move would be the best based on a
chess engine. This would guarantee victory. Would you use such a device to win?
What if an AI could predict exactly which colleges or majors would give you the most 70 57 44 88 64
money or happiness in the future? This AI would have the ability to simulate what
would happen in the future.
What if an AI could tell you whether someone would be a good friend or partner? 50 57 56 50 53
This AI would be able to tell you what will probably happen in the future.

Paradox of Choice Jelly Bean Case Study: Analysis of Jelly Bean Experiment:
1. Raw Data and Processed Data: Based on the above information, it can be concluded
that a smaller amount of choice leads to a larger amount
Table 2. The Effect of the Number of Choices of satisfaction because having two choices far exceeds
Available to Individuals on their Ratings of their having one choice or having five choices.
Satisfaction on a Scale of 1-10.
Trial # REFERENCES
# of Choices 1 2 3 4
1 5 3 4 5 [1] “How to Cheat at Chess.” Lichess.org,
2 5 7 7 lichess.org/how-to-cheat. Accessed 9 June 2021.
5 4 3 4
Table 3. Analysis of Table 2. [2] Schwartz, Barry (2004). “5”. The Paradox of
# of Choices Available Choice. New York, United States: Harper
Median Mean Range Perennial. ISBN 0-06-000568-8.
1 4.5 4.25 2
2 7 6.3 2 [3] Iyengar, S. S., & Lepper, M. R. (2000). When
5 4 3.7 1 choice is demotivating: Can one desire too much of
2. Graph: a good thing? Journal of Personality and Social
Psychology, 79(6), 995–1006.
https://doi.org/10.1037/0022-3514.79.6.995.
[4] Chernev, Alexander; Böckenholt, U.; Goodman,
J.K. (2015). “Choice overload: A conceptual
review and meta-analysis”. Journal of Consumer
Psychology. 25 (2): 333–358.
doi:10.1016/j.jcps.2014.08.002.
[5] Hill, Kashmir. “How Target Figured out a Teen
Girl Was Pregnant before Her Father Did.”
Forbes.com, 16 Feb. 2012,
www.forbes.com/sites/kashmirhill/2012/02/16/how
-target-figured-out-a-teen-girl-was-pregnant-
before-her-father-did/?sh=110e9e9b6668.
Figure 3 The Effect of the Number of Choices [6] DUMRONGKIAT MALA. “AI to Help Students
Available to Individuals on their Mean Ratings of their
Select Majors.” Bangkokpost.com, 27 Nov. 2018,
Satisfaction on a Scale of 1-10

1362
 Advances in Social Science, Education and Humanities Research, volume 631

www.bangkokpost.com/thailand/general/1582894/
ai-to-help-students-select-majors. Accessed 12
June 2021.
[7] Sakpal Manasi “Gartner Predicts 69% of Routine
Work Currently Done by Managers will Be Fully
Automated by 2024” 23 Jan 2020
https://www.gartner.com/en/newsroom/press-
releases/2020-01-23-gartner-predicts-69--of-
routine-work-currently-done-b.
[8] Baron-Cohen, Simon; Leslie, Alan M.; Frith, Uta
(October 1985). “Does the autistic child have a
“theory of mind”?”. Cognition. 21 (1): 37- 46.
doi:10.1016/0010-0277(85)90022-8. PMID
2934210. S2CID 14955234.Pdf.
[9] Wang, Zhiliang (2007). Smith, Michael J.;
Salvendy, Gavriel (eds.). “Artificial Psychology”.
Human Interface and the Management of
Information. Methods, Techniques and Tools in
Information Design. Lecture Notes in Computer
Science. Springer Berlin Heidelberg. 4557: 208–
217. doi:10.1007/978-3-540-73345-4_25. ISBN
9783540733454. S2CID 13060657.
[10] Winfield Alan F. T. (2018) Experiments in
Artificial Theory of Mind: From Safety to Story-
Telling Frontiers in Robotics and AI
https://www.frontiersin.org/article/10.3389/frobt.2
018.00075 10.3389/frobt.2018.00075 ISBN
2296-9144.

1363