ChapterPDF Available

The Social Impact of 3G, 4G, and 5G on Hong Kong from 2006-2020



This is the cover story in "Official Guide of ICT Industry in Hong Kong 2021", where my colleagues and I draw on social psychology literature and interviews with industry leaders to discuss the social impact of 3G, 4G, and 5G in Hong Kong from 2006-2020.
Cover Story
The Social Impact of 3G, 4G, and 5G on
Hong Kong from 2006-2020
3G4G 和 5G 從 2006-2020 年對香港構成的社會影響
By Cecilia Cheng, Ph.D., Linus Chan, M.A., Hubert Chan
Hong Kong is a world leader in the telecommunications industry.
Coupled with the amazing ability of Hong Kong‘s people to
embrace the latest and the best technology and services, the
Industry continues to spark innovations and novel applications
in the 5G era and make it a reality. Because of the advanced
connectivity of 5G, its social impact will be much more signicant
compared with the previous generations. This article is reviewing
the development journey starting from 3G and its respective
social impact all the way to 5G.
Overview of 3G (2006 - 2009)
The advent of the mobile phone was initially only capable of
calling (1G). Later, the introduction of 2G gave rise to shor t
message service (SMS). As the name implies, messages were
short since there was limited capacity for information exchange.
During 2006-2009, the advent of 3G increased the capability
of phones such that phones were no longer limited to calls and
texts. New affordances enabled 3G phones to play videos, use
apps, sur f the web, manage one’s calendar, and access one’s
social media accounts all on one device. Increases in data use
was driven by new applications such as the launch of Facebook
in 20 04, YouTube in 2005, and Twitter in 2006. By 2007, the
launch of the rst iPhone led to drastic increases with data use.
The iPhone was a key shift in the mobile phone industry as
it introduced and mainstreamed touch screens over physical
keypads. The iPhone can be used as a mini computer where
many apps are introduced. People could stream videos and
complete various tasks on the smartphone that require rapid
data transmission, which in turn prompted many operators
to find solutions to keep up with the demand. As sales of the
iPhone increased, so did the demand for using data via 3G. By
2008, the introduction of the Android gave operators alternate
platforms to disseminate their projects. Across both platforms,
there were clear demands for more data transmission.
To keep up with the demand, mobile op erators expanded its
multimedia services in 2006. Because watching videos on the
phone relies on the compatibility of the video format, not all
videos could be played and even then there may still be latency.
To address this issue, a breakthrough service (multimedia on the
move) was launched for phones to go beyond voice. In 2007,
public Wi-Fi service became more and more popular around HK.
In September 2007, there were more than 7,300 hotspots in more
than 4,000 locations around the city.
The increased demand in communications led CAHK to launch
the annual observance of World Telecommunication Day
(WTISD) in 2007, which marks the founding of the International
Te l e c o m m u n i c a t i o n U n i o n ( I T U ) o n M a y 1 7, 1 8 6 5 , d r a w i n g a t t e n t i o n
to the work of ITU and the challenges of global communication.
The purpose of this observance is to raise awareness of
the possibilities that the internet and other information and
communication technologies could bring to societies and
economies, as well as of ways to bridge the digital divide.
秀能力使電訊業能持續綻放光彩在 5G 時代裡推動更多的創新
科技與新興應並使之一一實現5G 先進的超卓連接力將會比以
顧由 3G 時代始起的發展歷程及其對 5G 時代帶來的社會影響。
3G 概覽 (2006 - 2009)
限,所 以 稱 之 訊。 在 2 0 0 6 - 2 0 0 9 年 間,第 三 代 動 通 訊( 3 G)
等。新 應 用 程 式 的 出 現 令 數 據 使 用 量 增 加,例 如 2 0 0 4 年 推 出 的 臉 書
(Facebook)2005 年的 YouTube 和 2006 年的推特 (Twitter)而在
2007 年面世的首部 iPhone 更令數據使用量大幅增加iPhone 可以
應用程式使 iPhone 可以當作微 型電腦般使用用家可以在其智能
iPhone 銷售額遞增透過 3G 的數據使用量亦同步增長到了2008
年,A n d r o i d 的 出 現 為電 訊 商 提 供 了另 一 個 選 擇 平 台去 發 布他 們 的 項
目。不 論 是 使 用 那 一 個 平 台,都 衍 生 出 對 數 輸 急 增 的 需 求 。
為滿足需求流動電訊商於 2006 年起開始擴展多媒體服務由於在
動多媒體)運而生令手機功能超越話音功能的限制。2007 年
個位置便設有 7,300 多個熱點
與日俱增的通訊需求亦驅使香港 通訊 業聯會於 2007 年起開始主
盟( I T U )於 1 8 6 5 5 1 7 日 成 而 設 目 的 是
Cover Story
With the principal support of the Office of Communications
Authority (OFCA) and a number of sponsors in the industry, CAHK
has been hosting the captioned event since 20 07, capitalizing on
the oppor tunity to promoting the latest in ICT to the public and
especially the youth and the academic communities. In this age
of technology revolution, young people are at the forefront, not
only the users but also as the innovators.
But how did initiatives like these alter how the populace’s lifestyle?
We arg ue that the o pen channels of soci al med ia pla tfor ms make
people more cautious of how they communicate online.
Smartphones and Novel Social Media
Contributes to Cautious Online Impression
The most notable changes in 3G are 1) the advent of new
affordances of smartphones where phones per formed like
portable computers, and 2) The ability to broadcast one’s life and
opinions via online photos, videos, and comments. Put together,
3G allows for accessible online interaction on their phones.
In conventional face to face interactions, people are motivated
to construct a favorable image to others. This idea stems from
impression management theory (Leary & Kowalski, 1990), which
proposes that people are motivated to construct a favorable
public impression that shows others only what is desirable (while
concealing what is undesirable). People consciously display
a favorable public self because it affects their status in the
social order with the ultimate goal of maximizing rewards and
minimizing punishments.
While impression management originated from people’s
interactions in the offline world, recent research suggests it is
also applied in the online world, where people are motivated to
portray a favorable public image to those they encounter on their
social networks. Online impression management mainly concerns
content-based management and network-based management
(Walther et al. 2008). Content-based per tains to the contents a
user places in the online world. As such, people carefully monitor
what is placed online to appeal to those they want to impress.
Network-based pertains to portraying one’s involvement in
certain groups, one’s social role within each group, and how one
communicates with others in that group. People’s sociometers –
a term in social psychology that refers to markers of self-esteem
based on the degree of inclusion and exclusion in social groups
– are broadcasted with public displays of friends, retweets, and
followers on various social media sites (Haidt & Rose-Stockwell,
2019). People can assess one’s social status to some degree
based on who one connects with online. In other words, who one
connects with on social networks influences how one's online
prole is perceived (Walther et al., 2008).
The 6th World Telecommunication and Information Society Day Ceremony - 2012
   
香港通訊業聯會自 2007 年開始主辦 WTISD藉此向大眾特別是
青少年與學界宣揚創新資訊科 技的最新資這個科 技革命的
代,走 在 最 前 線 的 青 少 年,不 再 是 科 技 的 使 用 者,更 是 科 的 創 新 者 。
3G 帶來最顯著的變化包括1) 智能電話的新屬性出現使手機可以
如同隨身電腦一樣操作2) 透過在網上發放圖片、視頻和評論來傳
播一個人的生活和觀點意見的能力。綜合來說3G 容許人們在其手
在日常面 面的交流 動中,會被 去建 一個良好 的形
這個概念來自印象管理理論(Leary 和 Kowalski, 1990 年)該理論
公眾自我形象,這是因為會影響他們在社會常規 秩序裡的地位而最
容的管理和基於網絡的管理(Walthe 等 2008 年)人們在網上世
儀 – 這是社會心理學上的一個術語意指基於在社交群組裡的融合
性和排他性的自尊標記 – 是從他們不同的社交媒體上公開展示的朋
轉發的貼文和關注者(Haidt Rose-Stockwell, 2019 年)
響人們如何看待他自己在網上的個人資料(Walther 等2008 年)
Cover Story
The inuence of social networks on the self has vast implications
for impression management. This is because the cyberspace
allows multiple users from different social groups to interact
simultaneously, and responses do not have to be synchronous.
Different social media accounts encourage people to highlight
different parts of them (e.g., LinkedIn vs. Facebook). Hence, the
audience across social media platforms is often conceptualized
as ‘imaginary’ since users cannot be certain if they are currently
being monitored and by how many people. Public social media
accounts (e.g., public Twitter and Instagram accounts) can be
viewed by anyone anonymously, which potentially allows people’s
online impressions to be under constant scrutiny. Since people
cannot immediately react and cater to the audience, they may
report feeling under constant surveillance.
Managing one’s impression with the ‘imaginary' audience is
difcult because the nature of connectivity of the Internet makes
a user connected to people from various social contexts (e.g.
colleagues, family, friends, acquaintances) simultaneously and
thereby collapses the context of different norms and different
social groups into one broad audience. This concept, known as
'context collapse,' forces people’s self-presentation strategies
to cater to various groups simultaneously. Groups that were
previously detached in the offline world come together in the
online world (Gil-Lopez et al., 2008). When one posts to a targeted
audience, they also have to consider that other recipients can view
the post. Since any online post can be seen by potentially anyone
in that particular social network, online impression management
tends to be more apprehensive as more social spheres collide.
Owing to context collapse, what people put online are broadcasted
to various groups simultaneously. Since people want to create
a favorable impression online, sometimes people publicly post
opinions, political views, and disagreements they do not personally
believe in to win the approval of others. These cases lead to people
publicly posting content online they do not actually believe privately,
known as preference falsication (Kuran, 1987). In other words, there
is high public endorsement for certain opinions, but little private
Preference falsication has implications on how people construct
their online persona. To cater to various groups at the same time,
people become hesitant to voice out their potentially controversial
views to avoid disapproval. To cater to more groups, people tend to
increase their linguistic variability and reveal personal information
about oneself only when it is universally appropriate. Marwick
and Boyd (2010) surveyed Twitter users who have hundreds of
thousands of followers and asked them who they tweet to, when
they self-censor, who they imagine reading their tweets, and
what makes them authentic. Results show that public tweets
mix the personal and professional, and are suited for a general
audience. Those Twitter users try to be authentic as best they can
without too much self-disclosure. Similarly, Gil-Lopez et al. (2008)
found that one’s self-presentation on Facebook is contingent on
one’s network size and network heterogeneity. The researchers
surveyed approximately 7000 Facebook users and found users
with both larger network size and larger network heterogeneity (i.e.
more Facebook friends from different groups) tend to have more
status updates but vary in whether posts are positive or negative.
Increased network size is associated with more positive status
updates, while increased heterogeneity is associated with more
negative status updates. The authors reason this latter nding by
asserting that those with a more diverse audience share more
negative emotions to garner social support. Alternatively, people
do not want to share too much positive news online if the audience
is diverse to avoid the perception of bragging.
面向(例如 LinkedIn vs. Facebook)。因,橫跨各種社交媒體平台
體的公開帳戶(例如公開的推特和 Instagram 帳戶)即潛在地容許
(context collapse)的概念迫使人們的自我展示策略同時滿足不
聚集在一起Gil-Lopez 等2008 年)當一個人向特定對象發布貼文
發布的貼文能 潛在地被任何人在特定的社 交網絡上查看故此隨著
這稱之為「偏好偽裝」(Kuran1987 年)換句話說大眾對這些意
「偏偽裝 」會影 響人們如 建他 網上 角色。為了同時迎
只在適當的情況下才會披露有關自己的個人訊息。Marwick 和 Boyd
(2010 年)對擁有數十萬關注者的推特用戶進行了調查詢問他
合,並 且 適 合 一 般 大 眾。那 些 推 特 用 戶 亦 會 在 不 披 露 太 多 個 人 訊 息 的
情況下盡力表現得最真實。同樣Gil-Lopez (2008 年)
的用戶(即較 多臉書朋友來自不同群組)往往 具有更多的狀態更
新,但 貼 文 屬 正 面 或 負 面 則 有 所 不 同。 網 絡 規 模 的 增 加 更 多 正 面 的
Cover Story
4G Launch
HKT announcement on 4G LTE 450Mbps launch in 2015. Mr. Peter
Lam, Managing Director of HKT’s Engineering (third right) and Mr. David
Wan g, P resi dent of Wirel ess Net work Pr oduc t L ine of Hua wei (th ird lef t)
demonstrate the world’s rst LTE-A 450Mbps solution in Hong Kong at
a press conference. They are accompanied by Dr. Henry Wong, Head
of Strategic Wireless Technology and Core Networks of HKT (second
right); Mr. SM Shuen, Head of Radio Networks Engineering and
Operation of HKT (far right); Mr. Bob Cai, President of LTE FDD Network
of Huawei (second left) and Mr. Xiao Yu, Deputy President, Product
Solution and Marketing, Southeast Asia Region, Huawei (far left).
   
  
  
Overview of 4G (2009 – 2019)
By 20 09, there was an intense need to keep up with the data
tsunami that resulted from smartphones. Consequently, many
initiatives within the Hong Kong telecommunications industry led
to enhanced speed and bandwidth to consume and disseminate
data. A significant event in 2014 was 4G Voice over Long Term
Evolution (LTE). The difference is twofold: rstly, the high density
voice makes the sound quality better than circuit-switched voice
(2G & 3G). Second, the efciency of the voice calls only takes 6
seconds to 8 seconds to connect the call. We highlight milestones
of some organizations:
HKT was aware that the phone was not used primarily for calls
anymore and from 2012 onwards phone calls became free.
Later, HKT had its own Public Cloud launch. HKT has bers in
Cheung Chau and Peng Chau. HKT also provided exclusive Wi-
Fi service in MTR line and MTR station. PCCW also had more
than 3,000 Wi-Fi hotspots.
China Mobile launched its 4G services with converged LTE
network in 2012 based on both FDD (Frequency Division Duplex)
and TDD (Time Division Duplex) standards. Based on the
statistics published by Quest Mobile in 2019, more than 80% of
China’s population is active mobile internet users now.
Hutchison noted that social media plays a huge role as
a transformation agent to facilitate information flow and
reshape social interaction between company and customers.
Facebook and Instagra m co ntinue to be two popular to ols for
companies to market new products and services, and interact
with customers. The emergence of social media tools has also
helped the company attract more young customers to subscribe
to the company’s services.
SmarTone noted that each network had to increase its data
capacity to meet customer demand which led to 4G. Since more
and more people use their phones beyond voice, SmarTone was
driven by a voice-driven to data-driven environment. It claimed
as the rst network in HK that supported 4G.
4G 概覽(2009 年至 2019 年
2009 年智能手機導致的數據海嘯引發迫切的數據處理需求 因此
帶寬 2014 年的重要發展是 4G 語音長期演進(LTE)而其帶來的
區別是雙重的首先高密度語音的音質優於電路交換語音(2G 和
3G)其次語音通話的效率僅需 6 到 8 秒即可連接我們在此重點
2012 年開始免費後來香港電訊推出自己的公共雲並且在長
洲和坪洲鋪設光纖更在港鐵線及港鐵站提供獨家 Wi-Fi 服務
電訊盈科當時還擁有三千多個 Wi-Fi 熱
中國移動於 2012 年推出了以 FDD(頻分雙工)和 T DD(時分雙工)
為基準融合的 LTE 網絡 4G 服務根據 Quest Mobile 在 2019 年發
布的統計數目前中國 80%以上的人口是活躍的流動互聯網用戶
之間的社交互動方面扮演著巨大的角色。臉書和 Instagram 仍然
SmarTone 指出每個網絡都必須增加其數據容以滿足客戶的
話,S m a r To n e 因 而 在 語 音驅 動 以 至 數 據 驅 動 的 環 境 下付 出 努力 。
SmarTone 聲稱是香港第一家支持 4G 的網絡公
Cover Story
4G Launch
3 Hong Kong launches 4G LTE network under
the “One Love. One World” banner in a large-
scale promotional campaign in 2012.
         
   
SmarTone has launched a territory-wide 4G
LT E n e t w o r k i n 2 0 1 2 , a n d w a s t h e o n l y m o b i l e
operator in Hong Kong support iPhone 5 with
4G LTE network.
           
   
   
Mr. Sean Lee, Director & Chief Executive
Ofcer of China Mobile Hong Kong.
Huawei capitalized on the potential
for 4G and have since employed over
80 thousand employees engaged in
Research & Development in scientific
research to ensure high-tech. Today,
they automated customer service and
upgraded their cloud service to allow for
massive data storage.
To w n g a s - a c o m p a n y w h i c h o r i g i n a l l y s p e c i a l i z e d i n t h e n a t u r a l
gas arena but later incorporated telecommunications business
– built optic fiber telecommunication tubes alongside their
gas tubes. Doing so provides uninterrupted optic fiber and
communication to data centers that provide cloud computing.
These data centers ensure confidentiality and safety from
cyberattacks and physical hazards (e.g. re).
TGT uses sophisticated Glass-In-Gas (GIG) technology to lay optical bres
within submarine gas pipeline, protected by polyethylene pipe, completing
an 1,800-meter cross harbor ber-optic path from Quarry Bay to Cha Kwo
Ling in February 2018.
 
    󲀛
  
華為充分利用了4G 的潛力自此
工,以 確 保 高 科 技 的 發 展如 今,
級了雲服務,以應 量的據存
香港中華煤氣有限公司 – 一家最初專注於天然氣領域,但後來
又合併了電信業務的公司。該公司在 燃氣管旁邊建 造了光纖
信。這 些 數 據 中 心 可 確 保 機 密 性 和 安 全 性,免 受 網 絡 攻 擊 和 人
Cover Story
To facilitate any potential disputes between providers and
customers, CAHK established the Customer Complaint
Settlement Scheme (CCSS) in 2012 to serve as a mediation
service. The CCSS is a mediation scheme set up by the
telecommunications industry resolves billing disputes in
deadlock between telecommunications service providers
and their customers. The mediation service is provided by
an independent service center (CCSS Centre) set up under
the Communications Association of Hong Kong (CAHK), an
industry association representing the communications sector
in Hong Kong. Customers who choose to use the mediation
service under the CCSS may first contact the Office of the
Communications Authority (OFCA) which will assess the
applications against prescribed criteria. OFCA will refer
accepted cases to the CCSS Centre for follow-up actions.
CAHK – [Customer Complaint Settlement Scheme] Grand Opening 2013
   
Meeting the increased data use has heavy social psychological
implications. We document six overarching trends:
1) Massive Increase in Data Use Contributes to Information
Overload and Fragmented Realities
2) Influx of Novel Social Media Platforms Altered Business-
Customer Communication and Led to the Rise of Key Opinion
3) Over-reliance on Phones Fuels Addiction, Polarization, and
Mental Health Concerns
4) Big Data is Capable of Constructing Online Personality Proles
and Processing Natural Language
5) Online Deception Provides a Safe Haven for Malicious Self-
Disclosure and Mass Misinformation
6) Ongoing Battle Between Cybercrimes and Cybersecurity
訊業聯會於 2012 年建立了解決顧客投訴計劃(CCSS作為調
服務CCSS 是電訊業建立的調解方案旨在協助電訊商與其客
戶解決一些已陷入僵局的計帳爭議 調解服務是由代表香港通訊
業界的組織香港通訊業聯會成立的代理機構(CCSS 中心)提供。
選擇使用 CCSS 調解服務的客戶可以先聯繫通訊事務管理局(通
交給 CCSS 中心採取後續行動
1) 數據使用的大量增加導致資訊超載和零碎的實況
2) 新興交媒體平台的蜂擁改變了企業與客戶之間的溝通導致關
3) 過度依賴手機助長成癮並導致兩極分化和心理健康問題
4) 大數據能夠構建網絡個性形象和處理自然語
5) 網上欺騙為惡意自我披露和大量錯誤資訊提供了避風港
6) 網絡罪案與網絡安全之間的持續對戰
Cover Story
Massive Increase in Data Use Contributes to
Information Overload, Crowdsourced Mobile
Apps, and Fragmented Realities
The massive increase in data use allows vast and rapid information
exchange that led to decentralized modes of creating and
sharing information, knowledge, and culture. With 4G, people can
download dozens of apps they could not previously because those
apps require high bandwidth that 3G did not provide. 4G also
allows people to document memories and events with photos and
videos, and people can choose to share these les with anyone
within minutes. Everyone can be an amateur reporter because
they can easily create and share content.
Everyone to some extent can choose who to share their online
content with (e.g. a video shared privately among friends vs. a
video posted on a public YouTube account) and due to context
collapse, public content tends to spread extremely rapidly.
The potential downside is that what is placed online potentially
stays online permanently, which hinders people’s control of
online impression management. An embarrassing video can
remain online for years to come. Any video, tweet, or comment
can be taken out of context and used against someone to libel,
shame, or ridicule. Worse, someone in possession of sensitive
information about someone else can use that information as
leverage for potential blackmail.
The spike of online content contributed to information overload
(Lee et al., 2017). To provide some order to information in the
cyberspace, interest groups are created for a particular niche.
For example, there are grou ps dedi cat ed for peop le interested in
a par ticular hobby (e.g. sports, music), groups for certain types
of people (e.g. expats, international teachers, people looking for
places to rent), and groups dedicated to discuss certain ideas
(e.g. politics, psychology, dating advice). Crowdsource apps were
created to facilitate discussion and communication for these
interest groups. Quora, for example, is a question-and-answer
platform founded by two former Facebook employees in 2010. A
user creates an account, asks questions under a topic, and wait for
other users to respond (Ovadia, 2011). Users can start a discussion
by comme nting, ranking, or de bating oth er users v ia their p osts.
Crowdsource mobile apps go beyond the virtual world and seep
into the physical world to the point it alters people’s lifestyles.
Rideshare apps such as Uber and Lyft altered the transportation
industry by crowdsourcing available drivers. Drivers for rideshare
apps are not professional drivers but rather citizens who can
drive. Lodging apps such as AirBnB altered the hotel industry by
crowdsourcing available accommodations. Hosts for lodging apps
do not run a professional hotel but are rather ordinary citizens
who have a place for others to stay temporarily. Dating apps
crowdsource singles looking to make new friends or potential new
partners. Other apps that crowdsource replace the need for in-
person interaction. Before 4G, it was common practice for people
in some countries to play chess with strangers in a public place.
With apps, people can play chess with strangers all over the world.
In a broader sense, crowdsource mobile apps allow people to
interact with similar others.
Despite the resources crowdsource apps offer, people still have
to selectively consume what websites to watch and what to
ignore due to information overload. Various Internet sites give
users control to customize their media consumption (e.g. which
news outlets to follow, which Twitter accounts to follow), where
most people choose ones that conform to their views (Riles et al.,
2018). This process of choosing what online content to consume
is not random, as people prefer to go after content that they
無法使用的應用程式因為這些應用程式需要的是 3G 無法提供的
高帶寬。4G 還允許人們使用照片和視頻記錄記憶和事件,人們可以
間私下共享的視頻對比在公共 YouTube 帳戶上發布的視頻)並且
可以利用這些 資訊 來進行潛在的勒索。
網上內容的激增導致 資訊超載(Lee 等2017 年)。為了向網絡空
。例如一些團體專門針對特定愛 (例如體育音樂)興趣的
人, 針 對 某 些 類 型 的 人( 例 如 外 國 人、 國 際 教 師 尋 找 租 房 的 人 )的 團
體,以 及 專 門 討 論 某 些 想 法 的 團 體( 例 如 政 治、心 理 學、約 會 建 議 )
Quora 是一個由兩位前臉書員工在 2010 年建立的問答平台由用戶
(Ov adia,2011 年)。用戶可以通 過其帖子拋 引玉向其他用 戶進
的生活方式。 諸如 Uber 和 Lyft 之類的共乘應用通過眾包可用的司
會開車的公諸如 AirBnB 之類的住宿應用通過眾包可用的住宿改
了面對面互動的需求在 4G 之某 些國家 / 地區的人們在公共場
特帳戶)大多數人可以在其中選擇符合其觀點的媒體(Riles 等
2018 年)。選擇要消費哪些網上內容的 程並不是隨機的,因為
人們傾向於追求他們同意的內。即使人們被動地上 馬遜、
Netflix 和 Instagram 也會根據用戶的網上活動向用戶建議他們可
Cover Story
agree with. Even when people passively surf the web, Amazon,
Netix, and Instagram would suggest users what they could buy,
watch, and whom to follow based on the user’s online activities.
Systematic ways of selecting what information to consume –
whether it be active through one’s volition, or passive through
algorithms – contributes to selective exposure.
Selective exposure is nothing new in terms of information
consumption. Even before the rise of telecommunications of the
Internet, people are bound to consume different information due
to various circumstances. That being said, selective exposure
is an issue in the 4G era as it creates a norm of fragmented
reality (Riles et al., 2018; Susskind, 2018). When people obtain
information about an event that they did not personally attend,
they turn to other sources of information such as viral photos/
videos, mainstream news media, blogs, or discussion boards.
While small-scale events such as conference talks or a viral
dance video can be captured accurately via online content (where
we define accuracy as consistent information across various
outlets and assume the information provided is true and not
manipulated), large scale events such as political movements
tend to provide lots of coverage that hold inconsistent information
across outlets. Some outlets report a certain political slant, while
another outlets contradicts that slant. People’s opinions of large
scale events are thus driven by selective exposure.
Information overload thus presents a puzzling paradox.
While more information gives people more access to various
perspectives, more information also means people consume
different perspectives, thereby causing people’s sense of reality
to be different than others when looking at the same event.
People’s reality is further fragmented with the influx of novel
social media platforms.
Inux of Novel Social Media Platforms Altered
Business-Customer Communication and Led to
the Rise of Key Opinion Leaders
Beyond fragmented realities, the 4G era coincided with the
promulgation of even more social media platforms that further
reshapes communication in the online world. Instagram and
Pinterest were launched in 2010, while WeChat and Snapchat
were launched in 2011. The emergence of these platforms
enhances connectivity in various ways. Generally speaking,
Instagram and Pinterest focus on broadcasting aesthetic
pictures to a wide range of people or organizations. WeChat and
Snapchat primarily serve as platforms to exchange photos and
videos in a more private manner.
These social media platforms influenced how some operators
communicated with their customers. They understood that social
media plays a huge role as a transformation agent to facilitate
information ow and reshape social interaction between company
and customers. As such, Facebook and Instagram continue to
be two popular tools for companies to market new products and
services, and to interact with customers. The emergence of social
media tools has also helped in attracting younger customers to
subscribe to the company’s services. Call centers took advantage
of social media to streamline their operations. Now, customers
no longer have to call to contact their customer service center.
Instead, customers can use WeChat, Whatsapp, and Facebook to
communicate. Doing so opens wide lines of communication that
were not previously possible.
These new social media platforms allowed everyone to be
connected around the world, with some emphasizing a public
form of connection. Instagram and Twitter users can “follow”
此,選 擇 性 曝 光 是 4 G 時 代 的 一 個 問 題, 因 為它 創 造 了 零 散 的 現 實 規
範( R i l e s 2 0 1 8 S u s s k i n d 2 0 1 8 年 )。當 人 們 得 他 們 自 己
瘋傳的照片 / 主流 新聞媒體博客或討論區。雖然可以通過網
通,並 導 致 主 要 意 見 領 袖(K O L )的 崛
除了零現實 外,4G 時 平台發表,從而
進一步重塑了網絡世界中的通信Instagram 和 Pinterest 於 2010
年推出而微信和 Snapchat 則於 2011 年推出這些平台的出現以
各種方式增強了連接性。一般來說Instagram 和 Pinterest 專注於
向廣泛的人或組織傳播美學圖片微信和 Snapchat 主要用作以更
到,社 交 媒 體 在 促 進 資 訊 和 重 塑 公 司 與 客 戶 之 間 的 社 交 互 動 方 面
扮演著巨大的角色。因此臉書和 Instagram 仍然是公司銷售新產品
了操作。如今,客戶不 再需 致電 繫其 客戶服務中心。而代,客
戶可以使用微信What sApp 和臉書進行交 流。樣做開通了以前無
繫,其 中 有 些 人 強 調 公 開 的 聯 繫 形 式。如 果 關 注 的 帳 戶 是 公 開 帳 戶,
有者(Ferrara 等2014 年)並且由於網絡效 人們可能傾向於
Cover Story
whomever they please, provided that the followed account is a
public one. One’s follower count serves as a marker for popularity.
More followers implicitly indicates that more people pay attention
to the owner of that par ticular account (Ferrara et al. 2014).
And because of the network effect, people may tend to trust an
account with more followers since it denotes more people trust it
(the same principle where people trust a restaurant that has more
customers than one that is empty). Of course, this asser tion that
followers equate popularity is contingent on the assumption that
each follower represents a distinct person/organization that is real.
It’s possible that some accounts are fake as some accounts are
fake or bots.
As people obtain more followers on their social media account,
their online posts (e.g. photos, videos, comments, blogs) are
broadcasted to more people. A post shared on an account with
a thousand followers would, on average, be seen more than a
post shared on an account with a hundred followers. Hence, an
account with more followers is more effective in communicating
with more people. When an account gets thousands of followers/
subscribers, the owner of the account is assumed to assert some
influence on the followers. Consequently, these people become
inuencers or key opinion leaders (KOLs).
KOLs rise in popularity in the online world due to an attractive
appearance, talent in a specific domain (e.g. singing, dancing),
their connection to a famous celebrity, or some aspect of their
personality. Because of the ‘Halo Effect’ a term in psychology
where people equate good-looking individuals with positive
traits – attractive KOLs are usually assumed to possess positive
personality traits (Harris & Bardney, 2019). KOLs attract and
retain their followers by frequently posting engaging content
that engages their viewers. Sometimes this requires preference
falsification, but nonetheless KOLs’ opinions influences their
followers. With enough popularity, it becomes common practice
for organizations ask KOLs to serve as the inter mediary between a
brand and customers. KOLs utilize the word-of-mouth advertising
technique but applied in a powerful online context due to context
collapse. The KOL economy is a rising trend in mainland China
and many businesses have turned to KOLs for advertising instead
of using conventional marketing channels.
Being a KOL and following KOLs stimulates heavy Internet Use.
As more people want to keep up with the latest trends, the latest
news, share their content with their fan base, and stay active in
communicating with businesses and customers, they end up
spending more time on their smartphones. Indeed, some research
found that people who thought about their smartphones were
more able to think of words that related to social networks, thereby
suggesting a connection between phones and social relationships
(Kardos et al., 2018). Smartphones provide opportunities for social
connections, but there comes a point where people over-rely on
their phones.
Over-Reliance on Phones Fuels Addiction,
Polarization, and Mental Health Concerns
A high-end smartphone in the 4G era enabled a plethora of new
affordances. Perhaps ironically, the smartphone is no longer used
much to make phone calls but instead used for other purposes in
a portable manner. People can store personal information, access
social networking accounts, travel, and store other personal les
all on their electronic devices. Users can access and “look-up”
all sor ts of information virtually anytime and anywhere because
time and geographical location are no longer constraints. People
can communicate with family, friends, associates, and colleagues
anytime regardless of where they are. Nowadays, the phone can
平均而言前者將被 多人看到。因此, 更多關 注者的帳戶在與更
多人 交 流 時 會 更 有 效 益。當一 個 帳 戶 有 成千上 萬 的 粉 絲 / 訂戶 時,
以假定該帳戶的擁有者對粉絲產生一定的影 響。因此,些人將成
KOL 由於具有吸引的外貌特定領域的才能(例如唱歌跳舞)
名人士的聯繫或其個性的某 些特質,KOL 在網絡世界中的知名度
此得以提升由於「光暈效應」(Halo Effect)- 這一心理學術語
KOL 具有積極的人格特(Harris 和 Bardney2019 年)。KOL 通
偽造偏好但 KOL 的意見仍然會影響其粉絲隨著 KOL 擁有足夠
的人氣機構要求 KO L 充當品牌 和客戶之間的中介已成為一種慣
由於社交媒體情境崩解KOL 利用口碑廣告技巧用在強大的網
上情境中。KOL 經濟在中國大陸呈上升趨許多企業已轉向 KOL
成為 KOL 和關注 KOL 會刺激互聯網的大量使用隨著越來越多的
人希望 上最 ,最新消 粉絲 群分享他們 內容,保持
詞,從 而 暗 示了 手 機 與 社 交 關 係 之 間 的 聯 繫( K a r d o s 等,2 0 1 8 年 )
4G 時代的高端智能手機帶來了量新功能。也許諷刺的是智能
電子設備上儲存個人資訊,取社交網絡帳 旅行以及儲存其他個
替了錢包的用途。 通卡已被 結到手機因此大家可以
Cover Story
also serve as a means for payment which largely substitutes the
wallet’s purpose. For instance, Octopus card is integrated into
one’s phone, so people can bring one less card and utilize the
phone as if it were an octopus card.
The drastic increase in data use is explained by many reasons,
one of which was the launch of several highly successful social
media platforms in a narrow window of time. Another reason is the
prevalence of smartphones. During the 4G era, the smartphone
market became saturated, which made everyone hyper-
connected all the time. While there are obvious benets such as
communicating with loved ones without time or geographical
constraints, hyper-connectivity have plenty of drawbacks to the
point some high-end restaurants do not want the disruption from
mobile phone activity and would forbid customers about phones
in the restaurants. As noted by one interviewee, hyper-connection
blurs the boundary between home and work, which makes it
difficult to ignore Email and texts during non-working hours.
Indeed, some experimental evidence found that checking Email
more increases stress (Kushlev & Dunn, 2015).
A sizable amount of research in psychology found that heavy
social media (and by extension Internet use) led to the rise of
Internet addiction disorder (Rosen, 2012; Weinstein et al., 2014).
Similar lines of research also found the habitual use of mobile apps
contribute to compulsive smartphone use (Clements & Boyle,
2018). Even the clinical psychology Diagnostic and Statistical
Manual, which classifies all known mental and health disorders,
included Internet gaming disorder in its current edition last
updated in 2013. Urges and behaviors regarding Internet use is
correlated with a whole host of psychological distress symptoms,
including insomnia, stress, compulsive need to check in with
technology, separation anxiety, and depression (Rosen, 2012;
Wang et al., 2018) . Active ly pay ing at tention to o ne’s smar tph one
while having face-to-face conversations is shown to be correlated
with relationship problems, negative workplace outcomes, and
conversation quality (Al-Saggaf & O'Donnell, 2019).
Additionally, heavy social media use contributes to polarization
as explained by three steps by a renowned social psychologist
Jonathan Haidt (Haidt & Rose-Stockwell, 2019). First, the
introduction of the Facebook “like” and the Twitter “retweet"
allowed content to be graded and ranked. Second, online posts
that contain humor or outrage are more likely to be shared
than mundane posts. Third, Facebook and Twitter present
one’s newsfeed for optimal engagement, such that posts with
the most likes, comments, and shares are presented instead
of chronological order. Taken together, social media platforms
facilitate the spread of outrage and this is exacerbated by
mainstream media. To increase news engagement, journalists on
social media platforms notice what is most popular and increase
viewership of their mainstream media by sharing the same popular
stories, which later gets shared back on social media. As a result,
stories that involve outrage receive more attention and have a
higher likelihood of going viral.
Sharing outrage could be both benevolent and malevolent. On
one hand, more transparency on the Internet shedding light on
unacceptable behavior (Farrow, 2019). On the other hand, sharing
outrage contributes to a “call-out” culture where any post can be
manipulated, taken out of context, or even fabricated to shame
and ridicule. As such, sensitive issues that benet from debate are
rarely communicated across parties publicly. Even acknowledging
a valid argument from someone in the opposing political/religious
party can easily get people in trouble from their own party, thereby
leaving little room for compromise (Murray, 2019). Instead, posts
about outrage are often rewarded and shared. Unfortunately, posts
that contain outrage tend to be uncivil, and sharing outrage on
出了多個 常成功 社交 平台。另一個原因是 能手 的普
擾,並 禁 止 顧 客 在 餐 廳 內 使 用 手 機。正 如 一 位 受 訪 者 所 指 出的 那 樣,
電子郵件和短。事實上一些實 證據發檢查電子郵件會增加
壓力(Kushlev 和 Dunn2015 年)
擴展使用導致了網絡成癮的上升(Rosen2012 年Weinstein 等
2014 年類似的研究也發現習慣性地使用流動應用程式會導致
強迫性使用智能手機Clements 和 Boyle2018 年)甚至對所有
在其 2013 年更新的最新版本中也包括互聯網游戲障礙與互聯網
迫切需要技術檢查,分離焦慮和抑鬱症(Rosen2012 年Wang
等,2 0 1 8 年 )。 面 對 時 積 注 意 自 己 的 能 手 機, 與 人 際 關
係問題負面工作場所結果和談話質量均有相關性(Al-Saggaf 和
O'Donnell2019 年
納森•海特(Jonathan Haidt)過三個步驟進行了解釋(Haidt 和
Rose-Stockwell2019 年首先臉書的「贊」和推特的「轉推」
進了憤怒的蔓 延,而主 媒體 加劇了這 憤怒。社交媒體平台注
分享憤怒既 是和 睦又 充滿 。一 網上的 更高 透明 揭露
人們無法接受的行為(Farrow2019 年)另一方面,共享憤怒會助
長「 召 喚 」文 化 ,在 這 文 化 中, 任 何 帖 子 以 被 縱, 斷 章 取 義 ,
各方之間公開交流。即使 認反 對的 黨 / 宗黨派中某人的有效
的餘地(Murray2019 年)相反有關憤怒的帖子通常會得到獎勵
較低(Popan 等2019 年)該研究表如果一方在網上表現出
Cover Story
the Internet exacerbates political polarization. Incivility on Internet
political discussions makes people perceive strong arguments
as irrational and aversive. Some researchers found that people
who encountered uncivil posts online rated the comments as less
rational regardless of argument is strong or weak (Popan et al.,
2019). This line of research implies that convincing others will not
work if one’s party behaves uncivil online.
Addiction and polarization to smartphones have been attributed to
mental health issues because of people’s over-reliance on them.
Although mobile communication can allow closer contact with
family members and reduce the psychological distance between
those living in different countries, people have less in-person
communication because we rely too much on social media tools
for communication by text. Also, people may be very anxious when
they lose their phone because it stores a lot of private information.
Losing one’s phone removes people’s access to online activities
such as escaping their real-life problems like loneliness.
On an affective (i.e. emotional based) level, the concerns behind
losing one’s phone seems to be widespread, such that the
expressions of ‘no mobile phone’ and ‘phobia’ has been combined
to form the psychological construct ‘nomophobia’, which refers
to separation anxiety for phones (Nie, Wang, & Lei, 2020). People
reported feeling anxious when thinking about their phone powering
off suddenly when engaging in online activities, such as messaging,
obtaining information, or recreation. Similar lines of research found
that those who score higher in smartphone separation anxiety tend
to have higher state anxiety, which in turn decreases their working
memory (i.e. the amount of information one can hold at a given time;
Hartano & Yang, 2016).
Additionally, people’s emotional states may plunge due to social
comparison. Smartphones allows people to connect with those
they would not otherwise meet in real-life. These include celebrities
and talented individuals all over the world. As such, people can
follow online proles of people who are superior or inferior to them
in a certain domain (e.g. singing, dancing, cooking, and acting)
for social comparison. When one compares to someone better,
they are conducting upward social comparison. But when one
compares to someone worse, they are conducting downward
social comparison. While upward social comparison can serve
as inspiration and gives people a goal to aim for (e.g. “I want to
be as good as Aaron Kwok at dancing”), more often it leads to
envy and decreases to subjective well-being and mental health
(Jang et al., 2017). In some cases, people want to self-enhance
their skills to compensate, and these cases of self-enhancement
are picked up by others (who then experience a fall in emotional
states), which in turn creates an envy circle (Verduyn et al., 2020).
Interestingly, some research found that social comparison leads
to a weaker decrease in depressive symptoms if people followed
fewer strangers on social media and instead followed more people
they knew (Lup et al., 2015).
But, this is not always the case if people experience the fear of
missing out. As people can see their friends broadcast their lives
on Facebook, Twitter, Instagram – some even on a daily basis – it
broadcasts many options that they themselves cannot be pursue
and thus ind uce a sense of exclusion. The form of exclusion
leads to the fear of missing out, which is the desire to continually
be par t of what other people are doing (Przybylski et al., 2013).
Some research found that those scoring high in fear of missing
out are associated with more depressive symptoms, less mindful
attention, and tend to check social media right after waking,
during meals, and even while operating motor vehicles (Baker et
al., 2016).
短了居住在不同國家 / 地區的人們之間的心理距但人們的面對面
樣,人 們 在 丟 失 手 機 時 可 能 會 非 常 擔 心,因 為 它 會 儲 存 很 多 私 人 資 訊 。
懼症」即手機的分離焦慮NieWang 和 Lei2020 年人們
低了他們的工作 記憶 (即人們 在特定的時間內可以保 存的資訊 量;
Hartano 和 Yang2016 年)
目標(例如「我想和郭富城一起跳 舞」更多時候它會引起嫉
並降低主觀幸福感和心理健康(Jang 等,2017 年)在某些情況下
人們想通 過自我提 升來彌補自己的能力而這些自我提升的案例
慕圈子(Verduyn 等, 2020 年)。有趣的是一些研究發現如果人
交比較而導致的抑鬱症狀會減弱(Lup 等2015 年)
Cover Story
On a behavioral level, some experimental evidence found that
the presence of smartphones decreases enjoyment on face-to-
face interactions (Dwyer et al., 2018). In their studies that sampled
hundreds of participants, people who were asked to keep their
phone on the dining table during a family meal felt more distracted
and enjoyed the meal less, relative to people who were asked to
put the phone away. Another study had participants maximize their
phone interruptions by turning phone alerts on for one week, then
minimizing their interruptions by turning notications off another
week. Participants reported higher inattention and hyperactivity
during the week notications were turned on, which in turn led to
lower productivity (Kushlev et al., 2016). Additionally, experimental
evidence found that smar tphone use decreases smiles between
strangers (Kushlev et al., 2019a). This line of research points to
the general finding that constant connectivity negates one’s full
attention and emotional benefits of social interaction in subtle
ways (Kushlev et al., 2019b).
Furthermore, social media promotes ever-present relational
aggression, which is dened as aggression based on manipulating
the victim’s relationships (Crick et al., 2002). Examples of relational
aggression online include threatening to unfriend/exclude, posting
embarrassing content to shame someone or to damage one’s
reputation by spreading rumors. Because people are hyper
connected, there is no escape from relational aggression. Even
when one blocks content from being posted on their profiles,
damaging content can still be posted online for everyone else to
see. Prolonged relational aggression has been associated with
increases in anxiety and depression, especially for young adults
(Twenge, 2017; Twenge & Martin, 2020).
On a cognitive (i.e. mental processes) level, the smartphone serves
as a source of transactive memory when one distributes the task
of encoding and remembering information from their own working
memory to their phone. That is, when people reduce the cognitive
demand of remember information (e.g. schedule, appointments,
addresses, passwords) by storing it in their phone, they no longer
have to remember it because the phone stores that information
(Risko & Gilbert, 2016). As a result, the phone serves as an
external memory source that people can consult with at any time.
While benecial in many ways, the downside is that the degree of
organization of one’s smartphone as a digital repository affects
people’s condence of recalling important information, such that
more organized digital repositories are associated with greater
condence in the knowledge stored within those systems (Hamilton
et al., 2016).
Constant connection to the Internet can backre as people tend to
mistake the Internet’s knowledge as their own (Ward et al., 2017).
Having instantaneously access to the Internet and the ability to
‘look up’ anything makes people believe they know more than
what they truly do, as evidenced by research that found people
believe answers they nd online are originally stored in their own
mind (Fisher et al., 2015). Instantaneous access to the Internet via
one’s smartphone can also lead to digital amnesia, which occurs
when people forget information that can easily be located on the
Internet (Wegner & Ward, 2013).
But letting go of one’s phone is not easy as the smartphone
is often a requirement whenever one leaves their home. This is
because the smartphone is capable of serving as one’s wallet
(e.g. digital payments such as Huawei pay, WeChat pay), one’s
GPS, and in some cases, one’s identication. In mainland China,
WeChat has evolved such that its 800+ million users can use it
to “hail a taxi, order food delivery, buy movie tickets, play casual
games, check in for a ight, send money to friends, access tness
tracker data, book a doctor appointment, get banking statements,
pay the water bill, find geo-targeted coupons, recognize music,
以看到他們 的朋 友在 書、推特、Ins tag ram 上 他們 生活,
的事情的一部分(Przybylski 等2013 年。一些研究發現那些害
起床後進餐時甚至在駕駛車輛時檢查社交媒體(Baker 等2016
互動的樂趣(Dwyer 2018 年)在對數百名參與者進行抽樣調
過度導致生產力降低(Kushlev等2016 年)此外實驗證據表明
使用智能手機可減少陌生人之間的笑容(Kushlev 等2019a
們對社交互動的全神貫注和情感收益(Kushlev 等2019b
害者關係的攻擊(Crick 等,2002 年)網上關係攻擊的例子包括威
脅要取消交友 / 排斥,發布令人尷尬的內容以羞辱某人或通 過散佈
Twe nge 和 M arti n,2 020 年)
自己的工作 分配到手 時,充當了交 互記 的來 源。也就
住它,因為電話會儲存該資訊(Risko 和 Gilbert2016 年)因此
人們回憶 要資訊的信心,更有組 織的數碼儲存庫讓人們
這些系統中儲存的知識更有信(Hamilton 等 ,2016 年)
的持續連接會適得其反(Ward 等2017 年。擁有立即存取互聯網
多,研 究 顯 示,人 們 認 為 自 己 在 網 上 找 到 的 答 最 初 是 儲 存 在 他 們 自
己的腦海中(Fisher 等2015 年)互聯網存取也會導致數碼失
(Wegner 和 Ward2013 年)
是必需的。 這是因為智能手機可以用作錢例如華為支付微信
支付等數碼付款)全球定位系 (GPS)以及某些情況下的身
Cover Story
search for a book at the local library, meet strangers . . . follow
celebrity news, read magazine articles, and even donate to charity
all on one platform” (Susskind, 2018).
4G has propelled an integration of people’s lives with smartphones
where people stay connected with others all the time. This is made
possible with data centers that ensure transmissions are quick and
safe. But, heavy use in smartphones contribute to Big Data that is
capable of predicting various human traits.
TGT Hong Kong Data Centre 2
 󱱣 
支付水費單、找地 理位 定向優惠券、識別音、在圖書館搜 索一
本書、結識陌生人 ... 關注名人新聞閱讀雜誌文章甚至捐贈給
善機構全都在一個平台上」(Susskind2018 年)
當人們處於一直與他人保持聯繫4G 便推動了生活與智能手機
Big Data is Capable of Constructing Online
Personality Proles and Processing Natural
Most social media platforms prompt users to upload some
identifiable information, such as a profile picture. Because the
prole picture often serves as the rst impression, people carefully
choose what to upload. But, the presence of profile pictures,
while useful to serve as the rst impression, can backre as those
pictures can be used to predict the user’s personality. Personality
research in the online world is conceptualized by the Five-Factor
Model, which claims to capture people’s personality on how
they vary on five overarching traits: openness to experience,
conscientiousness, extraversion, agreeableness, and neuroticism.
Qiu et al. (2015) recruited over five hundred Chinese participants
to complete personality scales, demographic information, and to
submit profile pictures voluntarily. Pictures were coded based on
various dimensions (e.g., camera position, face visibility, public
or private location) and these scores were correlated with the
self-reported personality traits. Results show that those picture
dimensions can predict basic demographic information (age,
gender) and some personality traits. Namely, the "duckface sele"
predicted neuroticism (i.e. sensitivity to negative emotions) and
youth, increased camera height predicted agreeableness (i.e.
compassion, politeness) and female, and private location predicted
conscientiousness (i.e. industriousness, orderliness) and male.
Internet users who learn about the predictive nature of profile
pictures may refrain from posting them, but people’s personalities
can still be predicted without pictures. Matz et al. (2017) aimed to
predict the personality traits of 3.7 million people to see how they
react to Facebook ads (measured by clicks and download rates).
They obtained people’s levels of extraversion from myPersonality.
org, where participants voluntarily completed personality scales
(again based on the Personality Five-Factor Model), and people's
Facebook likes. After building a data-driven model that predicts
levels of extraversion from Facebook likes, Matz et al. (2017) used
片。由於 用 戶 照 片 通 常 是 第 一印 象,因 此 人 們 謹 慎 選 擇 要 上 傳 的 內 容。
片可用於 測用戶的個性。網絡世界中的人格 究是 由五因素模型
神經質。Qiu 等(2015 年)招募了五百多名中國參與者以完成人格
圖片可是即使沒有照片人物的個性仍然可以預測Matz 等(2017
年)的目的是 測 370 萬人的個 性特 徵,以了解他們對臉書廣告的反
應( 通 過 點 擊 和 下 載 率 衡 量 )。 他 們 從 m y P e r s o n a l i t y . o r g 獲 得 了人
五因素模型並獲得了他們的臉書贊。Matz 等(2017 年立了
二。研 究 人 員 隨 後 為 這 些 臉 書 用 戶投 放 了七 天 的 廣 告,發 現 如 果 廣 告
擊並下載廣(即填字遊戲)1.3 倍。
Cover Story
an experimental paradigm that dichotomously classified person
extraversion (high vs. low) and ad extraversion (high vs. low). The
researchers then ran ads over seven days for those Facebook users
and found that ad engagement was higher if the ad matched the
extraversion of the Facebook users; people were more likely to click
and download the ad (i.e. a crossword puzzle game) 1.3 times.
This form of psychological targeting does not stop there. Even
without personality self-reports or profile pictures, people’s
digital footprints inadvertently reveal their personality traits and
demographic information. Since the advent of Facebook, the
Facebook ‘like’ became the most generic digital footprint that
reveals positive associations with products, sports, musicians,
books, brands, and other personal preferences. Kosinski et al. (2013)
obtained Facebook Likes (on photos, status updates, Facebook
pages) from 58,000 volunteers to predict their age, gender, sexual
orientation, ethnicity, religious and political views, personality
traits, happiness, and use of addictive substances. These outcome
variables were categorized dichotomously, and the predictions
were highly accurate. Computers were able to predict demographic
information with data-driven models with success rates of over
80% in some categories. These findings imply that people may
choose not to reveal certain aspects of oneself such as their
age or use of addictive substances, and yet this information can
be predicted from what they ‘like’ on Facebook. Given the ever-
increasing amounts of digital footprints people leave behind (and
some of which are placed by other people in one’s social network),
it becomes increasingly difcult for individuals to fully control what
attributes they reveal in the online world.
In fact, there is some evidence that computers are now better
at predicting one’s personality traits than one’s own Facebook
friends. Wu et al. (2015) recruited over 80,000 participants to
complete a 100-item personality questionnaire (i.e. International
personality item pool), and computers were tasked to predict the
personality of these participants by using only the participants’
Facebook 'likes'. Results show that the computer’s predictions
of personality via Facebook Likes were more predictive than
the user ’s Facebook friends’ rating of the user’s personality.
Subsequent analyses found that more likes from a user increase
prediction accuracy substantially, such that computers need
approximately 100 ‘likes' to outperform a friend and 300 ‘likes' to
outperform a spouse.
With abundant data about people’s online activities, computers
can conduct advanced correlational and Bayesian analyses to
predict its users’ behaviors. Indeed, one interviewee noted that
there are intelligent searches when one buys something from
Ta o B a o o r A m a z o n . T h e s e i n t e l l i g e n t s e a r c h e s w o r k b y a n a l y z i n g
what people tend to buy. If many people who purchase item X tend
to also purchase item Y (thereby indicating a positive correlation
between item X and item Y), then people who buy item X are
notified about item Y. As more data is obtained, more complex
algorithms are able to predict what cluster of items people will
likely buy.
As it stands now, Big Data is capable of predicting demographic
information and behavioral patterns. But it does not stop there;
computers are able to process people’s language. Psycholinguists
and psychologists have long found that people's choice of words
inadvertently reveals demographic information, personality traits,
and thought processes (Pinker, 2007). As such, psychologists are
interested in studying people’s language online, which gave rise to
computerized textual analysis. In the past, studying how language
reveals one’s emotionality, social relationships, deception, thinking
styles, and demographic information has been cumbersome and
limited because researchers have to code responses by hand and
much of the research participants were college students (Tausczik
聯繫Kosinski 等(2013 年從 58,000 名志願者那裡獲得了臉書贊
向、種 族、宗 教 和 政 見、人 格 特 質、幸 福 和成 癮 性 物 品 的 使 用。將 這 些
模型預測某些類別的人口統計資訊成功率超過 80%這些發現
用,但 是 這 些 資 訊 可 以通 過 他 們 在 臉 書上 的「贊 」來 預 測。 鑑 於 人們
日益留下數碼足 (其中一些足 被其他 放置在某人的社交網
中)人們 完全 自己網絡 界中所展 屬性。
格特徵 Wu 等(2015 年)招募了 80,000 多名參與者以完成一項
100 個項目的個性問卷(即國際個性項目庫)並要求電腦僅使用參
與者的臉書「贊」預測這些參與者的個性 結果顯示電腦通過
性。隨 後 的 分 析 發 現,來自 用 戶 的 更 多「 贊 」大 大 提 高了 預 測 準 確 性,
因此電腦需要大約 100 個「贊」來勝過朋和需要 30 0 個「贊」
人們購買東西的傾向來操作。果許多購買商品 X 的人也傾向於購
買商品 Y(從而表明商品 X 和商品 Y 之間存在正相關關係)那麼就
會向購買商品 X 的人通知有關商品 Y。隨著獲得越來越多的數據便
和思維過程(Pinker2007 年)因此心理學家對網上學習人們的
與者都是大學生(Tausczik 和 Pennebaker2010 年 )自然
目前探索與字詞計算(Linguistic Inquiry Word Count)是用於研究
Mehl 和 Niederhoffer2003 年)由於其原有概念是研究負面情緒
寫作的有益效果探索 與字詞計算現在可以根據 70 多種文本形音義
(例如代詞、 知過 程、知過程、道德 行為)蹤個人敘話、
相關結果(Pennebaker2013 年)。例如,在整個生命週期中人們
第一人稱代詞表示狀態低落,鬱症 狀以及誠
& Pennebaker, 2010). Today, research methodology behind natural
language processing has changed drastically by collecting and
analyzing big samples of text across a variety of domains in a
rapid fashion.
Currently, one of the most widely used programs to study what
words reveal about people psychologically is the Linguistic Inquiry
Word Count (Pennebaker, Mehl, & Niederhoffer, 2003). Since its
original conception that studies the beneficial effects of writing
about negative emotions, the Linguistic Inquiry Word Count now
tracks personal narratives, conversations, novels, and other genres
based on over 70 textual dimensions (e.g., pronouns, cognitive
processes, perceptual processes, moral foundations). While
counting words may sound rudimentary, counting words alone
is already capable of predicting people’s age, status, deception,
health, and related outcomes (Pennebaker, 2013). For instance,
people use more positive words and fewer first-person singular,
more past tense, and more complex words over the lifespan.
Using rst-person pronouns is indicative of low status, depressive
symptoms, and also honesty.
Apart from the Linguistic Inquiry Word Count, there are other
programs that process people's language to analyze their psyche.
One recently-created computer program analyses the complexity
of texts (Conway et al., 2014). This computer program called the
Automated Integrative Complexity (or Auto IC) scores the overall
complexity of a passage of text based on differentiation (number
of separate concepts evoked about a particular topic) and
integration (connections made between the topics evoked). Simple
texts indicate a relatively unidimensional structure of an idea,
while more complex texts represent multi-faceted thoughts about
an idea. Further, the Auto Integrative Complexity system scores
elaborative complexity (the extent to which the text is one-sided)
and dialectical complexity (the extent to which the text considers
alternative perspectives).
The complexity of ideas has shown to predict a plethora of
significant outcomes. For example, complexity predicts war
outbreaks, maintaining political power, lying, health, and terrorism
(Repke et al., 2018; Conway & Woodard, 2020). Leaders are
capable of manipulating how complex (or how simple) they
communicate as a means of inuence (Chan et al. 2020). Clearly,
computers can use digital footprints to understand people based
on linguistic content and complexity, and these programs to
analyze data are much quicker than human coders.
Every time someone uses the Internet for impression management
strategies to convey a specific persona or to search something
on a search engine, they leave digital footprints that let computers
learn more about them. Even what we listen to on Spotify, watch
on Netflix, or order on Amazon becomes data for computers to
learn about us (Mitnick, 2017). In other words, our digital footprints
reveal more than what we want others to know. Given this
revelation, are people motivated to hide their true selves online?
In the next section, we argue why online deception cultivates both
malicious self-disclosure and mass misinformation.
Online Deception Provides a Safe Haven
for Malicious Self-Disclosure and Mass
There are psychological mechanisms that facilitate deception in
the online world. We discuss the most prominent one: the online
disinhibition effect. People often behave differently in the online
world because psychological restraints that regulate behavior
are weakened, which in turn disinhibits people's behavior. In one
of the most widely cited cyberpsychology articles, Suler (2004)
Cover Story
除了探 索 與 字詞 計 算外,還有 其他 程 式 可以 處 理人 們 的語 言以 分 析其
心理最 近創建的一種電腦程式可以分析文本的複雜性(Conway 等
2014 年該電腦程式稱為自動化綜合複雜性(或「自動 IC」
爆發維持政治權力撒謊,健康和恐怖主義(Repke 等2018 年
Conway 和 Woodard2020 年領導者有能力操縱他們溝通的複
雜程度(或簡單程度)作為一種影響力的手(Chan 等2020 年)
關他們的資訊甚至我們在 Spotify 上聆聽 N etflix 上觀看或在亞
馬遜上訂購的內容都變成了電腦了解我們的數據(Mitn ick2017 年)
有一些心 機制 可以造成 絡世界中的欺騙。們現論最 突出
引用的網絡心理學文章之一中Suler(2004 年)解釋了這種現象背
explains the six mechanisms behind the phenomenon known as
the online disinhibition effect. The rst mechanism is dissociative
anonymity, which argues that the online self is compartmentalized
since anonymity detaches one's online self from one’s ofine self.
The second mechanism is invisibility, which argues that since
people cannot see one another online (even when they know
the other ’s identity), they are less concerned about how they
sound or look when they communicate. The third mechanism
is asynchrony, which means people do not interact in real-time.
Delays in response time could lead to people coming back to
the conversation whenever one feels it is convenient to do so,
occasionally to the point of what Suler calls ‘emotional hit-and-
run.’ The fourth mechanism is solipsistic introjection, which
occurs when people subvocalize as they read their texts as if they
are talking to themselves. Disinhibition strengthens when this
introjected character feels subjectively real. The fth mechanism
is dissociative imagination, which occurs when people feel that
their online personas are characters they created in the online
realm is in an entirely separate space. The last mechanism is the
minimization of status and authority, which argues that norms in
the online world are less strict and less enforced. Taken together,
these six factors explain why people tend to behave more
disinhibited online, and recent research supports its relevance to
the online world today (Schlosser, 2020).
Malicious Self-Disclosure
Online disinhibition has implications for developing and maintaining
interpersonal relationships in the online world. Namely, the degree
to which people reveal personal information about oneself –
known as self-disclosure - signals trust, liking, and increased
commitment to the relationship (Reis & Shaver, 1988; Sprecher,
et al. 2013). Recipients can interpret disclosure as desire to form
a more intimate relationship, which in turn causes people to
disclose to those whom they like and trust since self-disclosure is
simultaneously vulnerable and rewarding (Tamir & Mitchell, 2012).
Individuals who disclose intimate topics report increased
closeness relative to non-intimate/non-emotional ‘small-talk’
topics. Notably, disclosure can be categorized as factual (e.g.
descriptive statements) or emotional (e.g. private feelings and
judgments), with the latter being a stronger predictor of intimacy
than self-disclosure of facts.
Although self-disclosure has various benefits, it can be risky as
revealing personal information can be used against oneself if there
is a record of what is said in the online world. To this end, some
people might fake self-disclosure to get others to self-disclose to
obtain sensitive information that can be used for future blackmail.
Mass Misinformation
Online deception is 1) intended to mislead and 2) communicated in
a technologically mediated message (Hancock, 20 07). In the online
world, deception is broadly classied as either message-based or
identity-based. As the name implies, message-based deception
refers to creating or sharing fictitious messages. To qualify as a
deceptive message, the sender must be aware that the message
is fabricated while the receiver is not aware (Finneman & Thomas,
2018). There are many forms of message-based deception,
including full fabrication, fact fabrication, date-time fabrication,
source fabrication, quote fabrication, and plagiarism.
But is message-based deception a unique issue in the online
world? Some research found that there is more message
deception when people communicate online. For instance,
Naquin et al. (2010) argued that Email communication is subject
to more negative language, more inhibited behavior, and is more
Cover Story
Suler 所說的「情緒奔跑 (emotional hit-and-run) 的地步。第四種
時,就 會 發 生 這 種 分 離。最 後 一 種 機 制 是 地 位和 權 威 的 最 小 化,該 機
歡和增強對關係的承諾(Reis 和 Shaver1988 年Sprecher 等
2013 年接受者可以將披露解釋為渴望建立更親密的關係,這又
(Tamir 和 Mitchell2 012 年)
與非親密 / 非情感化的「閒聊主題相比披露親密主題的人表示親
露,而 讓 其 他 人 進 行 自我 披 露,以 獲 取 將 來 可 用 於 勒 索 的 敏 感 資 訊。
網絡欺騙是 1)意圖誤導和 2)以技術為媒介的訊息交流(Hanc ock
2007 年)。 在網絡世界,欺騙大致分為基於訊 息的或基於身份
的。顧 名 思 義,基 於 訊 息 的 欺 騙 是 指創 造 或 分 享虛 擬 訊 息。要 獲 得 欺
(Finnem an 和 Thomas,2018 年)。基於訊息的欺騙 有很多形式
究發現當人們網上交流時會有更多的訊息欺騙。例Naquin 等
(2010 年)認為電子郵件交流更容易受到負面語言的影響行為受
psychologically distant. Participants in their study were randomly
assigned to play the ultimatum game (i.e. a game about distributing
resources between two people) where players communicated
via Email or pen-and-paper. Results show that those in the Email
condition deceived more, and this effect was explained by the
permissiveness of misrepresenting their answers online.
In contrast, identity-based deception pertains to creating a
false identity. In the online world, anyone can lie about personal
information (e.g., age, gender, ethnicity, occupation, life
experiences) and can join or leave virtually any online context
whenever they please. Because the majority of social networking
sites do not require identity verication or pose a restriction on how
many accounts one can possess, little regulation stops identity
deception. Without regulation, instances of malicious identity
deception are rather prevalent in the cyberspace. For instance,
there are established groups that specialize in online romance
scams. Whitty and Buchanan (2015) examined the psychological
impacts of online romance scams by interviewing victims. In most
cases, the scammers gain the victim’s trust by self-disclosing
and declaring one’s love and the intention of starting an exclusive
relationship with the victim, followed by frequent and intense
communications. The scammer will then ask for gifts that increase
in monetary value over time. When the victim nds out the online
romance is a scam, some reported that the emotional loss was
more damaging than the nancial loss, some had issues coping,
and many had negative views of online relationships afterward.
Instances like these where scammers manipulate victims to
disclose sensitive information is known as social engineering.
Social engineering methods often capitalize on both message
and identity deception to achieve their goals of embarrassment,
blackmail, and manipulation (Hancock, 2007; Whitty & Young,
2017; Mitnick, 2017). A lucid example would be phishing scams,
where scammers successfully tricked victims into disclosing their
Email passwords because scammers pretended to be authority
gures to get victims to disclose personal information.
Ye t d e s p i t e t h e p o t e n t i a l l y d e t r i m e n t a l e f f e c t s o f o n l i n e d e c e p t i o n ,
there are arguments for concealing one’s identity. People conceal
some facets of their identity because the online world is a safe
haven to avoid perceived stigma behind marginalized identities
that are biologically based (e.g., ethnic or sexual minorities) or
choice-based (e.g., drug users). Alternatively, some may choose
to conceal their identity because they treat the online world as
a form of escape that provides anonymity they cannot attain in
the ofine world. However, the outcomes of concealment are not
always clear. Some research found that endorsement to conceal
one’s marginalized identity was positively associated with self-
esteem (Plante et al., 2013), but other research nds that long-term
concealment of marginalized identities is associated with anxiety
and depression (Pachankis, 2007). Regardless of the outcome,
there are clear motivations for online deception.
Social engineers have clear motivations behind online deception,
and advances in digital technology continue to expand their arsenal.
Technology today that manipulates fake photos and videos to
sound or look like someone else is known as “Deepfakes.” Deepfake
technology arose from computers learning to replicate patterns
from Big Data, and applications of this technology are beginning
to proliferate. Although this sort of technology can be beneficial
in some ways such as recreating the voice of a historical figure
to educate children, or to restore speech to people who lost their
voice, social engineers can employ Deepfakes to put people’s faces
on unsettling avenues (e.g., adult videos) to be used to blackmail,
to intimidate, or to plant fake evidence (Chesney & Citron, 2019).
Implications of Deepfakes are huge, as fact and fiction become
harder to discern. As audio and video content cannot always be
Cover Story
和 Buchanan(2015 年通過採訪受害者考察了網絡浪漫騙局的心
交工程方法通常利用訊 息和身份欺騙 來實現其尷尬、索和操縱的
目的( Ha nc oc k2 00 7 年Wh it t y 和 Yo un g2 017 年M it n ic k2 017 年)
電子郵件密因為詐騙者 假裝是權威人物以誘使 受害 者披 露個
是,隱 藏 的 結 果 並不 總 是 很 清 楚。一 些 研 究 發現,認 可 隱 瞞自 己 的 邊
緣化身份與自尊成正相關(Plante 等2013 年,但其他研究發現
長期隱瞞邊緣 化身份與 慮和抑 相關(Pachankis2007 年)
起來像其他人的這種技術被稱為「深偽 (Deepfakes)。Deepfake
社會工程師可以利用 Deepfakes 將人們的臉龐放在令不安的途
徑( 例 如, 人 視 頻 )用 於 勒 索 、 恐 嚇 或 植 入 虛 假 證 據( C h e s n e y
Citron2019 年)Deepfakes 的含義是巨大的因為事實與虛構變
taken at face value, fraudulent videos become harder to discredit,
and authentic videos might, in turn, become more easily discredited.
Consequently, misinformation is exacerbated.
The prevalence of online deception cultivates mass misinformation.
While the exact rate of misinformation online is challenging to
determine, people do not know what to trust or not trust. As a result,
there have been calls for governments and tech companies to work
together to increase cybersecurity and reduce misinformation (Smith
& Browne, 2019). However, increasing cybersecurity is no easy task
due the increased prociency of cybercrimes.
Ongoing Battle Between Cybercrimes and
Advances in smartphone technology inevitably come with it the
risk of cybercrimes. In today's digital age, countless individuals
use the Internet for work and leisure on a daily basis, but the
downside is that increased information technology use predicts
increased susceptibility to cybercrime victimization (Cheng, Chan,
& Chau, 2020). Unfortunately, implementing effective cybersecurity
is strenuous because it has to defend against the unique
characteristics of cybercrime (Chan & Chan, 2018b). Namely,
cybercrimes can 1) attack regardless of physical location, 2) attack
at any time, 3) be anonymous and untraceable in some cases, 4)
succeed with only one attempt. In some ways, cybercrimes, which
are continually evolving alongside technological advances, are
more dangerous than conventional crimes.
Some of the most common types of cybercrime according to the
Cyber Securit y and Technology Crime Bureau of the Hong Kong
Police Force are: online game-rel ated c rime, Inter net /e-mail a ccount
abuse, e-mail scams, hacking, online shopping fraud, other online
fraud (e.g., e-auction fraud), credit card misuse, and social media
deception. Further, technological-based cybercrimes include
viruses, spyware, ransomware, and distributed denial of service
(DDoS). As cybersecurity measures increase, so too do cybercrime
methodologies, thereby creating a mutual battle between
cybercrime and cybersecurity. Given the prevalence of cybercrime in
Hong Kong – where the total number of reported cybercrime cases
in Hong Kong has soared more than vefold from 2009 to 2018 - we
can be reassured that the operators have implemented measures
for their customer’s data security. China Mobile encrypts its
customer data which makes individual information to be impossibly
interpreted by intruders. Huawei has implemented ngerprint and
face ID lock that secures customers’ data.
Our interviewees noted that the ICT sector are working arduously
to enhance cybersecurity. Different organizations in the world such
as European Union and international standard bodies continue to
study the 5G Cybersecurity concern, and mitigation measures will
be deployed to minimize the security risk. Another way to keep
your mobile data safe is to store them in cloud servers instead of
storing them in your phone. But no matter how secure one’s data
is, there is still a non-zero chance that people fall to cybercrime. In
fact, cybercrimes tend to target new applications that might yield
the most valuable data to exploit, whether it be a new social media
app or a video conferencing platform.
What happens after cybercrime occurs? Some research found
that falling prey to cybercrimes – specifically the most common
cybercrimes listed by the Hong Kong police force mentioned
above - is associated with decreased perceived fairness, control,
trust in the online world, life satisfaction, and happiness (Cheng
et al., 2020). With ever-present social engineering ploys, people
have to remain vigilant in all online activities to prevent those
detrimental outcomes from occurring. Given that what is posted
Cover Story
能性會越來越大(ChengChan 和 Chau2020 年)。不幸的是
(Ch an 和 Chan,2018 b)即網 絡罪 案可以1)不受物理位置的攻
2)隨時攻3)在某 些情 況下是匿名且不可追踪的;4)僅一次嘗
型包括與網絡遊戲有關的罪案互聯網 / 電子郵件帳戶濫用電子
安全之間引發一場相互對戰鑑於香港的網絡罪案盛行(從 2009
年到 2018 年 )香港網絡罪案舉報案例總數激增逾五倍。但我們可
我們的受訪者指出ICT 業界正在努力提高網絡安全性世界各地的
不同組織如歐盟和國際標準組都在持續研究 5G 網絡安全
零。事 實 上,無 論 是 新 的 社 交 媒 體 應 用 程 式 還 是 視 頻 會 議 平 台,網絡
控制力對網絡世界的信任活滿意度和幸福感有關(Cheng 等
2020 年)隨著社會工程手段的不斷發展人們必須在所有網上活動
尬的照片都可能被用於勒索和操縱(Mitnick,2017 年)如果其他用
online can remain permanent, any embarrassing photo posted
on social media platforms or related arenas by oneself or others
could potentially be used for blackmail and manipulation (Mitnick,
2017). Even Facebook posts or Tweets that are deleted can still
be exploited if other users saved the post before deletion. In sum,
the new affordances provided by smartphones in the 4G era
opens doors for cybercrimes, which in turn call for increases in
The era of 4G prompted many changes in the human psyche;
some for the better and some for the worse. We next move onto
the social psychological impact of 5G.
Social Psychological Impact of 5G (2020 onwards)
The year 2020 marks the launch of the new 5G era that will
transform industries and societies to fully integrate automated
immersive technology. 5G would increase data transmission
(including upload and download speeds), connect millions of
smart devices, and reduce latency. Soon, applications behind
augmented reality, artificial intelligence, virtual reality, and the
Internet of Things would make Hong Kong to become a smart
city. But, getting there would require initiatives from the Hong
Kong telecom operators and related research institutions and
Achieving proper handover to 5G mobile networks would result in
uninterrupted service while developing new applications. These
include developing Cloud computing, AI, cloud games, AR, and
VR. 5G can boost the operational efciency, new service offerings,
and enhance customer experience of the enterprises. The
telecommunication industry facilitates the ow of people, goods,
capital and information which can drive economic development
of a society. In time, 5G would support fast nancial transactions
and bring more innovation to the nancial industry such as smart
banking, smart nance and asset management.
Cover Story
可以被利。總而言之,智能手機在 4G 時代提供的新能力為網絡
4G 時代引發了人類心理的許多變化有些變得更好有些變得更糟
接下來我們將探討 5G 的社會心理影響。
5G 帶來的社會心理影響(2020 年以後
2020 年標誌著嶄新的 5G 時代的來臨它將改變各行各業及社會
將之全面融入自動化的科技年代5G 將增加數據傳(包括上載及
目標需要香港電訊營運相關的研究機 和製 造商通力合作,
在開發新應用程式時,需要正確切換至 5G 流動網絡才不會令服務
出現中斷,這些包括雲計算AI雲遊戲、AR 和 VR。5G 能夠提高營
亦支援快 速金融交易為金融業帶來更多創新的應用如智慧銀
Ms. Susanna Hui, Group Managing Director of HKT (center), together with
(from left) Mr. Peter Lam, Managing Director of HKT Engineering; Mr. Tom
Chan, Managing Director of HKT’s Commercial Group; Mr. Bruce Lam,
Managing Director of CSL Consumer Mobile, announce 5G service to be
launched on April 1.
     
Guest-of-Honour Miss Eliza Lee Man-ching, Permanent Secretary for
Commerce and Economic Development (Communications and Creative
Industries) (2nd from left), Dr. Li Feng, Chairman of CMHK (2nd from right),
Mr. Sean Lee, Director & Chief Executive Ofcer of CMHK (1st from left),
and Dr. Max Ma, Director and Executive Vice President of CMHK (1st
from right), jointly ofciated at the Grand Opening of the China Mobile 5G
Innovation Centre Hong Kong Open Lab.
   
  
Upcoming Trends in 5G and its
Psychological Impact
The upcoming years will be met with
upcoming trends that capitalize 5G
technology. These include 1) increased use of
public and private data networks to maximize
efciency, 2) new ways of interacting with the
world through augmented reality and vir tual
reality, 3) connecting millions of devices
through the Internet of Things with ultra-
low latency, and 4) computers complete
human tasks via articial intelligence. Taken
together, these trends will make Hong Kong
into a smart city, which will radically enhance
various institutions.
Increased Use of Public Data to Maximize
The pros and cons of using public data have long revolved around
the conflict between security and privacy. While some trust the
government or other tech organizations to monitor the public’s
activities (e.g. surveillance, facial identification, traffic data) for
purposes of safety, accountability, and efficiency, others feel
that any monitoring of them and their movements is an invasion
of privacy. Indeed, public data may sometimes be anonymous
in principle but Big Data is already capable of predicting one’s
personality and behaviors based on their digital footprints.
Identifying people in security footage naturally means that some
level of information about an individual or a group is somewhere
in cyberspace. While surveillance and facial recognition has
positive benets such as deterring crime and preventing identity
theft, facial recognition could still be problematic for a few
reasons. For one, security breaches are an ever-present danger
in our increasingly digital world (Marr, 2019). Surveillance and
facial identication software places the security of an individual's
HTHKH Executive Director and CEO Kenny Koo (third from left) and his
management team announce 3 Hong Kong’s launch of 5G service.
    
SmarTone launched its world-class 5G service in May 2020, with the city’s
widest network coverage through Dynamic Spectrum Sharing (DSS) to
spearhead smart city development in Hong Kong .
       
 
Cover Story
5G 的未來趨勢及其帶來的心理影響
未來數年5G 科技的發展趨勢將會帶來好處
別,自 然 地 意 味 著 可 以在 網 絡 空 間,取 得 關 於 個 人 或 團 體 一 定 程 度 的
年)。監視及 部識 軟件把個人資料的安全置在任 何能
Cover Story
personal information in the hands of whatever entity collected
that information, rather than with the individual, thereby taking
away a level of personal control that many people are attached
to. For another, if people do not know the intent of a particular
corporation or entity that collects information via facial recognition,
they might worry that the corporation will use that information for
purposes they may not agree with, such as targeted marketing.
A third reason is that public data compromises one’s capacity for
impression management in the online world. As discussed in a
previous section, people are motivated to craft a favorable image
(especially KOLs) by choosing what they want others to see (and
what to conceal). Facial recognition and surveillance potentially
bring information to light that inuencers never intended to share.
For example, a renowned vegan health blogger would not want
to be seen indulging in a Big Mac. This debate between privacy
and security is echoed by our interviewees. For instance, lamp
posts that monitors trafc and road conditions have been met with
concerns on both privacy and security. While these lampposts can
accurately report any collisions/road accidents, it can also capture
what people are doing in vehicles. As a result, PCCW decreased
the resolution of their cameras to protect people’s privacy while
maintaining a much-needed record of trafc conditions.
New Ways of Interacting with the World through
Augmented Reality and Virtual Reality
Another trend in the 5G era is seeing the world mediated by
technological advances. One way technology alters our perception
is through augmented reality, which places computer-generated
stimuli that appear on one’s smartphone. Today, augmented reality
is already used in real-time translation (e.g. Google Translate app),
shopping and marketing (McLean & Wilson, 2019), education
(Akçayır & Akçayır, 2017), and video games (Morschheuser
et al., 2017). As more apps that feature augmented reality
increase, people will be able to see a different reality through
their smar tphones. The possibilities are endless are computer-
generated images on reality will be available for smartphone users.
SmarTone pushes the frontier to 5G, pioneers Hong Kong’s rst 5G live eld trial in 3.5 GHz and 28
GHz bands simultaneously, also showcased the future Mixed Reality for town planning.
後,可 能 會 進 行 一 些 未 經 他 們 同 意 的 用 途,例 如 進 行 目 標 市 場 推 廣 。
前一節的論述,人們被激 勵去 構一個良好的形象尤其是 KOL
受關注即使這些街燈可以準確地匯報任何碰撞 / 路面意外但同時
5G 時代的另一個趨勢是藉由先進科技作為媒介去了解這個世
即時 翻譯(如Google 翻 、購物、(McLean 和
Wilson2019 年)教育AkçayırAkçayır,2 0 17 年 )和 視 頻 遊 戲
(Morschheuser 等,2017 年)。隨著具有擴增實境功能的應用程
China Mobile NB-IOT Launch
 
Another way technology alters our perception is through virtual
reality, which takes over what we see and hear. Current applications
of VR involve video games, education, and art. For instance, the
PlayStation 4 has games that allow players to immerse themselves
in a haunted house or a battleeld. Medical students can use VR
applications such as “Share care” to visualizes various human
organs. These organs can be viewed in 3D space that no traditional
textbook can provide. Artists can use apps such as “Tiltbrush”
to paint in a variety of colors, brush styles, and in a 3D setting
without using physical canvases (Chan & Chan, 2018a). As one
interviewee pointed out, VR allows anyone to ‘see’ popular sights
around the world. People can sightsee world wonders (e.g. Eiffel
To w e r , t h e P y r a m i d s ) w i t h o u t b e i n g p h y s i c a l l y p r e s e n t . H u t c h i s o n
anticipates that VR and AR technologies will provide consumers
with a seamless experience with hotel check-ins and services, as
well as virtual room experience and customization.
Connecting Millions of Devices through the
Internet of Things
These technologies alter how we see the world are made possible
with scalability and network management 5G provides by making
these devices more powerful when devices are connected via
the Internet of Things (Li et al., 2018). 5G-IoT allows collaboration
between smart devices to mutually interact and share data without
human input. In time, people could communicate in a highly
interactive and immersive virtual world with low latency to simulate
reality. Conferences, classrooms, and meetings can be done
entirely virtually.
However, the inherent diverse nature of networks and devices in
5G-IoT renders a lack of consistency across IOT systems and
applications. Hence, there needs to be technology standards (i.e.
data aggregation, network protocols) and regulatory standards
(i.e. security solutions). For instance, HKT completed successful
outdoor 5G trials in both the 3.5 GHz and 28GHz spectrum bands
and first in Hong Kong to do a live outdoor broadcast on 5G
network in 2018. China Mobile collaborated with their partners
to provide more services like transmission, broadband, IoT and
Information and Communications Technology (ICT) solutions.
Cover Story
方式現時 VR 的應用範圍有視頻遊戲教育和藝術層面等例如
PlayStation 4 備有利用 VR 的遊 戲讓玩家恍如置身在鬼屋或戰
醫科學生可以利用「Share care」等 VR 應用程式來將人體不同
看這些器官術家可以運用「Tilt Brush」等應用程式在 3D 設置
(Chan 和 Chan, 2018a)正如其中一位受訪者所述VR 讓任何人
賞世界奇(如艾菲爾鐵塔和金字塔)和記電訊預VR 和 AR
5G 提供的可擴展性和網絡管理令終端裝置通過物聯網連接後的性
可能(Li 等2018 年)5G-IoT 容許智能終端裝置之間進行互動
不過5G-IoT 中的網絡固有多樣化特性和終端裝置致使 IoT 系統
測試並率先在香港使用 5G 網絡進行實時室外廣播中國移動香港
亦與合作夥伴提供更多服務如傳頻服務IoT 和資訊通訊科
技( I C T )解 等 。
Cover Story
Computers Complete Human Tasks via Articial
As technology continues to advance, computers will be able to
outperform many humans in various tasks. Nowadays, computers
can possess similar cognitive functions as humans to learn and
achieve their goals, albeit in specific domains (Tegmark, 2018).
Computers learn via machine learning to build schemas/models
by sifting throu gh large datase ts. For i nstance, com puters learn to
recognize objects and faces by going through heaps of pictures.
Computers learn to compose music by sifting through heaps
of music. Computers learn how to read emotions (i.e., affective
computing, sentiment analysis) by linking various stimuli with
typical human reactions.
Applications of AI would signicantly alter nance, manufacturing,
education, transpor tation, energy, healthcare, legal systems, and
other industries (Tegmark, 2018). Because AI is still in its relative
infancy and the long-term outcomes of implementing AI in various
industries are still largely unknown, questions as to whether AI is
an overall benet or an overall danger to humanity is controversial.
While AI is capable of advancing many industries, integrating
computers into human lives has large-scale implications that
can potentially backre. For instance, while computers can help
humans make better decisions with data-driven models derived
from Big Data (Denis & Young, 2017), computers can also collect
data on humans that can be used to predict them. Additionally,
those who grow up with AI technology may anthropomorphize
life-like technology and attribute them with human-like psyches
and thereby inuence children’s social and cognitive development
(Severson & Carlson, 2010). Some evidence suggests that children
attribute digital assistants with psychological properties but
no biological properties as a new ontological category of living
things. This line of research found that children would talk to
digital assistants like a friend and condant for social and moral
deliberations (Severson & Woodard, 2018).
Even the best researchers in Articial Intelligence do not have a
consensus on whether AI will surpass human competence or gain
digital consciousness to think on its own. What AI researchers do
agree on is that computers surpass humans on specic tasks (e.g.,
playing chess), but not all tasks in general (Tegmark, 2018). Since
AI technology utilizes machine learning and big data to automate
tasks that are typically done by humans (with the unfortunate
outcome of putting many people out of jobs), MIT professor Max
Tegmark recommends today’s youth to aim for careers where
machines are bad at. These jobs include heavy interpersonal
communication, unpredictability, and creativity.
However, communication tasks do not necessarily give humans an
advantage. Computers are learning to outperform humans on tasks
such as writing. Using data mining to extract how people write in
various genres, programs such as Grammarly not only understand
spelling and grammar, but can also improve writing based on
criteria such as audience, formality, domain (e.g. academic,
general, technical), tone, and intent (e.g. describe, convince,
inform). Grammarly can check for plagiarism across millions of
written texts online and even provide automated feedback that is
now implemented in businesses and universities (O’Neill & Russell,
2019). However, it does not stop there. With machine learning,
computers can already learn to write journalistic pieces (e.g.,
stock prices, business report) like humans in arenas that rapidly
converts data into narrative news entirely without human input.
This practice of automated journalism publishes articles at a much
faster pace than most journalists with the added ability to create
multiple customized versions for different audiences (based on
their digital footprints). Articles that are primarily fact-based such
目標(Tegmark2018 年)。電腦藉著機器學,通過篩 選大 數據
集來建立模式 / 模型例如電腦藉助大量的圖片去學習識別物體和
AI 應用將會為金融、製造教育通運輸能源、醫療保建法律
系統及其他行業帶來顯著改變(Tegmark2018 年)由於 AI 仍處
於起步階段並且在各個行業中運用 AI 的長期結果仍是未知之數
因此有關 AI 對人類是有整體利益還是有害仍然受到爭議。然 AI
出更好的決定(Denis 和 Young2017 年,但電腦同樣可以從人類
身上收 得來 對其作出測。此外,那些 AI 科 一起 成長
從而影響孩子的社交和認知發展(Severson 和 Carlson2010 年)
碼助理像朋友和知己般傾談以進行社會和道德思Severson 和
Woodard2018 年)
即使是人工智能領域裡最優秀的研究人員也無法就 AI 是否能超
識。A I 的 研 究 人 員 同 意 在 特 定 任 務 上( 如 下 棋 ),電 腦 能 超 越 人 類,
卻非所有任務(Tegmark2018 年。由於 AI 技術是通過機器學習
令到很多人因此而失去工作)故此麻省理工學院(MIT)授 Max
Teg mark 建 現今的年青人,要找一些機 不擅長的,包括一些
在各種體裁中的寫作方式例如 Grammarly 之類的程式不單能
作。G r a m m a r l y 亦 能 在 網 上 過 百 萬 的 文 章 裡 檢 查 有 否 出 現 抄 襲,亦
Russell2019 年)但是它並不止於此,電腦通過機器學習已可
責編寫並預計這種新聞形式將會激增(Carlson2015 年
The 3Innocity programme supports start-ups,
industry leaders and scientic research talent to
build a digital economy ecosystem.
  
as spor ts, finance, and marketing have been
written by computers (without human input) for
several years now, and this form of journalism is
expected to proliferate (Car lson, 2015).
Articial intelligence analyzes heaps of data to
learn how to write as competently or, in some
cases, better than humans. Computers will
continue to improve as more data is fed into
the online world. As computers become more
capable, it will learn to perform and enhance
various industries. As more and more industr ies
becoming ‘smar t’, Hong Kong will become a
smart city.
Hutchison 5G Base Station
  
Making Hong Kong a Smart City: Enhancing
Transportation, Healthcare, Education, Shopping,
and Entertainment
When there is lots of data that is fed into millions of connected
devices that lets people see the world in a technologically
mediated way, then over time Hong Kong would become a smart
city. Smart cities as a concept include ICT infrastructures that are
increasingly digitized, inter-connected, and sustainable, such that
there is smart transportation, smart healthcare, smart education,
smart shopping, smar t entertainment, and smar t industries all
around. Admittedly, we are still at the beginning of 5G so some
applications below are theoretical.
Cover Story
Smart Transportation
Currently, there are apps (e.g. rideshare apps, Google maps) that
inform people how to get from location to destination based on the
transportation available. These apps inform people where certain
vehicles are at and approximately when it will reach a destination.
However, these apps focus on getting people from point A to
point B and overlook other factors that might affect large-scale
transportation. To enhance transportation, there needs to be lots of
data transmitted among vehicles in real-time with ultra-low latency
to maximize efficiency and reduce collisions. Doing so includes
real-time location tracking, sensors to reduce collision between
vehicles, maximizing on-demand transport, instant notications of
delays and collisions, and to fully utilize parking spaces.
Traffic data can be gathered in real-time via GPS devices, road
cameras, vehicle identification, and even crowdsourced traffic
reports (Amin-Naser i et al., 2018). Information collected involves
location and speed of vehicles and traffic conditions, with the
hopes of creating predictive algorithms to inform the public when
travelling is most efcient. Trafc data would be disseminated via
radio, texts, digitized trafc signs, or apps to inform citizens about
optimal routes.
人工智能通過分析大量 數據
秀。當 越 來 越 多 的 數 據 進 入 網
當大量數據流經數百萬個相互連接的終端裝 置時人們便會在科技
智能購物智能娛樂和智能工業隨處可見。誠然我們仍處於 5G
現時,有些應用程式(如共乘應用程式Google 地圖)會提供如何
從一個位 置到達另一個目的地的交通選這些應用程式會顯示附
指示人們如何從 A 點到達 B 點而忽略了其他可能引致大型交通
化、延 遲 和 碰 撞 即 時 通 報 和 充 分 利 用 停 泊 位 置 。
實時交通數據可以通過 GPS 終端裝置道路攝影車輛識別
至群眾外包的交通報告來獲取(A min -Naseri 等2018 年)。收集的