Paradigm Shift in Human-Machine Interaction: A New Learning

Framework for Required Competencies in the Age of

Artificial Intelligence?

Michael Burkhard, Sabine Seufert and Josef Guggemos

Institute for Educational Management and Technologies, University of St. Gallen,

St. Jakob-Strasse 21, 9000 St. Gallen, Switzerland

Keywords: Smart Machine, Artificial Intelligence, Human Augmentation, Education, Learning Framework.

Abstract: Smart machines (e.g., chatbots, social robots) are increasingly able to perform cognitive tasks and become

more compatible with us. What are the implications of this new situation for the competency requirements in

the 21st century? This paper evaluates the underlying paradigm shift with relation to smart machines in edu-

cation. It discusses the potentials and current limitations of smart machines in education in order to eliminate

prejudices and to contribute to a more comprehensive picture of the technological advances. In light of human

augmentation, the paper further proposes a possible learning framework that includes the human-smart ma-

chine relationship as a normative orientation for new competency requirements.

1 INTRODUCTION

The 21

century confronts us with a variety of chal-

lenges. Globalization, increasing digitalization and a

longer life expectancy in industrialized countries are

changing our work and our lives. The conventional

three-stage model of education, work and retirement

is increasingly changing into a multistage life (Grat-

ton & Scott, 2016).

A technology that has made particularly great

progress in recent years is artificial intelligence (AI).

AI includes different elements such as the ability to

solve domain-independent problems and the ability to

interact and learn from its environment (Dellermann,

Ebel, Söllner & Leimeister, 2019, p. 638). In form of

smart machines, AI can, to a certain extent, make

decisions and solve problems without the help of a

human being (Pereira, 2019). Chatbots (e.g.,

Amazon´s Alexa) or social robots (e.g., SoftBank

Robotics´ Model Pepper) can be considered as

manifestations of such smart machines. Today, smart

machines support us in everyday life, but have also

the potential to replace us (Wike & Stokes, 2018). For

example, in the field of natural language processing,

GPT-3, an autoregressive language model with 175

billion parameters, is capable of generating news

articles that human reviewers can hardly distinguish

from articles written by humans (Brown et al., 2020).

A number of critical voices are being raised,

indicating that we may need to redefine our role as

humans in relation to smart machines (e.g.,

Davenport & Kirby, 2016; Floridi, 2016; Aoun, 2017;

Jarrahi, 2018; Baldwin & Forslid, 2020).

With respect to education, this may imply that we

have to question our competencies and strengths and

redefine them in relation to smart machines. There are

tasks that smart machines can do better than we can.

Aoun (2017, p. 53) requests “a new model of learning

that enables learners to understand the highly

technological world around them” and calls his model

humanics, with the goal of providing a “robot-proof”

education (Aoun, 2017, pp. 53-61). In addition to a

“robot-proof” knowledge, competencies that help us

to collaborate with smart machines in everyday life

could be valuable. Suto (2013, p. 139) introduces the

concept of robot literacy, a media literacy focusing

on forming appropriate relationships with smart

machines. Since our image of smart machines is often

influenced by Hollywood movies such as I,Robot or

Ex machina, it is sometimes difficult to keep an

unbiased, neutral picture of the technology. Digital

natives are not always familiar with new technologies

and may need to be actively made aware of and

informed about these technologies (Ng, 2012). Even

though certain students may be able to acquire digital

skills automatically by growing up as digital natives

294

Burkhard, M., Seufert, S. and Guggemos, J.

Paradigm Shift in Human-Machine Interaction: A New Learning Framework for Required Competencies in the Age of Artiﬁcial Intelligence?.

DOI: 10.5220/0010473302940302

In Proceedings of the 13th International Conference on Computer Supported Education (CSEDU 2021) - Volume 2, pages 294-302

ISBN: 978-989-758-502-9

in the internet age, it may be fruitful to actively foster

the digital competencies of students in the sense of a

skill-based perspective (List, 2019).

Jarrahi (2018, pp. 578-579) argues that the

discussion about AI is often characterized by two

positions: one side claiming that smart machines “will

soon outthink humans and replace them in the

workplace”, the other side assumes that the concern

around AI is just “another overhyped proposition”

(Jarrahi, 2018, p. 578). According to Jarrahi (2018),

it would be better to think about ways, how humans

and smart machines could combine their individual

strengths in a synergetic way. The augmentation of

human skills is central (Davenport & Kirby, 2016). In

the future, those people who work at the interface

with smart machines and are able to blend their skills,

might be in high demand (Quick, 2019).

Existing learning frameworks, such as the OECD

learning framework 2030, recognize the challenges

associated with AI and call for new solutions in a

rapidly changing world (Organisation for Economic

Co-operation and Development [OECD], 2018, p. 3).

To prepare for 2030, students as the decision makers

of tomorrow should have a broad set of competencies,

consisting of knowledge, skills, attitudes and values

(OECD, 2018, p. 5). Further, students should be

competent in creating new value, reconciling tensions

and dilemmas and taking responsibility (OECD,

2018, p. 5). While we agree that students need such

competencies, we believe that it may be useful to

evaluate the role and competencies of the student in

relation to smart machines.

As smart machines can act autonomously and take

over tasks for the students, smart machines may

become more than just an advanced tool. We argue

that smart machines increasingly take over a co-roll

towards us, which has implications on our

competency requirements. Smart machines may

become cooperation and learning partners. In light of

the identified research desideratum, the following

research question should be addressed:

How do smart machines influence the competency

requirements for the 21

century?

The objectives of the paper at hand are therefore

twofold:

 Evaluating the underlying paradigm shift of smart

machines with regard to education, in order to bet-

ter understand current developments, obtain a

more unbiased picture and evaluate underlying as-

sumptions;

 Explaining the implications and raising awareness

for smart machines with regard to competency re-

quirements, with the goal to contribute to a nor-

mative foundation for the use of the technology.

Our paper can serve as a starting point for future

research, as it highlights important concepts and

variables related to competency requirements with

regard to smart machines and provides a normative

orientation. In the light of human augmentation, we

want to contribute to a more comprehensive picture

of the technology in order to eliminate prejudices and

to lay the foundation for better decisions on the use of

smart machines.

To this end, we lay the foundation for the

emerging competency requirements in section 2. We

take a closer look at the driving forces of change,

which result in the merging of smart machines with

our everyday lives and discuss the potentials and

current limitations of smart machines. Section 3

explains the potential paradigm shift in education

with regard to smart machines and makes a

proposition towards an extended learning framework.

Section 4 concludes with some final remarks.

2 CHANGING LEARNING AND

WORKING CONDITIONS

2.1 Driving Forces of Change

The competency requirements in the 21st century de-

pend on various developments. Among other things,

our society is challenged by demographic change, in-

creasing digitalization, and globalization.

Life expectancy in most countries has risen

sharply during the last century. At some point in time

a 100-year life could be possible (Gratton & Scott,

2016). The life phase of work is increasingly charac-

terized by a large number of stages combined with re-

orientation. The conventional three-stage model of

education, work and retirement is increasingly chang-

ing into a multistage life (Gratton & Scott, 2016).

Therefore, the ability to engage in life-long learning

is becoming increasingly important.

At the same time, the increasing digitization also

presents us with new challenges. Connecting a wide

range of devices to the internet enables us to generate

more data than ever before (Floridi, 2013, p. 5). Deal-

ing with data and the associated problem solving is

becoming increasingly important and is regarded as

an important 21

century skill (Aoun, 2017; Rios et

al., 2020). Due to the “half-life of facts” (Arbesman,

2013), it is difficult to stay up to date. Advances in AI

Paradigm Shift in Human-Machine Interaction: A New Learning Framework for Required Competencies in the Age of Artiﬁcial

Intelligence?

295

are leading to increasingly smart machines that help

us recognizing patterns in data and deal with the in-

formation overload. However, AI technology also

carries the risk of partially replacing us, as smart ma-

chines are able to perform and automate cognitive

tasks.

In the wake of globalization, this tendency is fur-

ther intensified. Baldwin and Forslid (2020) speak of

a twin trend of globalization and robotics (globotics).

Successful business models are spreading around the

world and arbitrage opportunities provide financial

incentives to drive globalization further (Baldwin,

2019, p. 63). Regulatory attempts by individual coun-

tries have only a limited effect. It is the tech compa-

nies like Google who are creating new realities (e.g.,

in the field of autonomous driving)

(https://waymo.com/journey).

2.2 Smart Machines Merge with Our

Everyday Life

According to Floridi (2013, pp. 3-5) we are in transi-

tion to a new era of hyperhistory in which we will be-

come increasingly dependent on our own technologi-

cal achievements. Information and communication

technologies (ICT) are not only used to record and

transmit data, but also to process it increasingly au-

tonomously. Information as a fundamental resource is

becoming increasingly important and life offline is

hard to imagine. Unlike the era of history, people be-

come dependent on ICT in the era of hyperhistory.

Although most people today still live historically,

there is a shift to a hyperhistorical life (Floridi, 2016).

We are in the process of creating an infosphere, an

informational environment comparable to, but differ-

ent from, cyber space (Floridi, 2013, p. 6). The crea-

tion of an infosphere shared by human and digital

agents is facilitated by the fact that machines are be-

coming increasingly smart. To a certain extent, they

are becoming more compatible with us humans and

can take over more areas of our expertise.

Smart machines are able to perceive and interact

with their environment because they have the ability

to process large unstructured amounts of data due to

the underlying technology of AI. As Figure 1 shows,

this process of perception, reasoning and action is

currently facing the two obstacles of black box and

data bias.

First, smart machines are currently often a black

box who are difficult or impossible to interpret (Zor-

noza, 2020). New “explainable AI”-approaches would

be important to understand the reasoning of the smart

machine (Zornoza, 2020). As long as smart machines

are not explainable, they cannot be fully trusted, which

restricts the current use of the technology.

Second, since every smart machine is only as

good as the data on which it was trained, biased train-

ing data (e.g., with regard to gender, race, religion,

etc.) can lead to stereotyped or prejudiced content

generated by smart machines (Brown et al., 2020, p.

36). As some of the best smart machines today, are

mostly trained with data from the internet (Brown et

al., 2020, p. 9), these fears are justified and addressed

by researchers (Brown et. al., 2020, pp. 36-39).

Smart machine may currently be restricted in their

use due to technological limitations. However, al-

ready today, smart machines can improve organiza-

tional decision making (Jarrahi, 2018). Raisamo et al.

(2019) argues that smart machines can help humans

to augment their capabilities. Similar to eyeglasses,

smart machines could act as a tool to compensate for

possible weaknesses and enhance strengths (Raisamo

et al., 2019, p. 131). Different researchers are also ex-

ploring how smart machines can support us not only

Source: Adapted from SBFI (2019, p. 30).

Figure 1: Capabilities of Smart Machines to interact with the Environment.

InteractionwithEnvironment/Human

Perception

Images,speech,

text,sensordata

Clustering,dimensionreduction,classification,regression,...

Applications

Principles

Reasoning

Problemsolving,

planning,diagnosis,

decisionmaking

Action

Speechoutput,

recommendations,

robotics/navigation

Functions

Data‐ Learning‐ Predictions

Bias

Black

Box

CSEDU 2021 - 13th International Conference on Computer Supported Education

296

as a tool, but as a partner. For example, Dellermann

et al. (2019) introduce the concept of hybrid intelli-

gence, which combines the complementary intelli-

gence of humans and smart machines. Together, the

human and the smart machine “create a socio-techno-

logical ensemble that is able to overcome the current

limitations of (artificial) intelligence” (Dellermann et

al., 2019, p. 640). As Dellermann et al. (2019, p. 640)

point out, smart machines and humans learn from

each other through experience and improve over time.

Wesche and Sonderegger (2019) go even one step

further and argue that smart machines could in the fu-

ture increasingly act as a leader, as smart machines

have begun to take over leadership functions by guid-

ing and commanding human workers. Figure 2 illus-

trates the potential evolution of smart machines and

its transformative effect on society.

For now, many smart machines often serve as tool

to support us in a controlled, static and programmed

environment as for example the free translation tool

DeepL (https://www.deepl.com). Even though these

smart machines are already very useful, the trans-

formative effect on society is comparatively small.

However, smarter machines step into a co-role with

us, becoming increasingly autonomous, self-learning

and independent, the greater are the transformative

effects and the associated challenges for society.

Whether we as society live in a world where smart

machines evolve into a co-role to us is also a norma-

tive question that needs to be discussed. In certain ar-

eas (e.g., autonomous driving), many people may be

willing to let smart machines take the lead. In other

areas (e.g., in law enforcement), it will probably still

be of critical importance that humans have the last

word. No matter which path we will choose as soci-

ety, it may be valuable to anticipate the current devel-

opments and search for solutions, how humans and

smart machines can act in a complementary and mu-

tually reinforcing way.

3 SMART MACHINES AS

LEARNING PARTNERS?

3.1 Paradigm Shift of Mutual

Dependency

For education, the upcoming of smart machines means

that we have to question our own strengths and compe-

tencies and redefine them in relation to smart ma-

chines. The internet provides us with a large amount of

current data and information that can be accessed by

students at any time. The threshold between offline and

online is becoming blurred. Today, we live an “onlife”

life in a global infosphere (Floridi 2013, p. 8). Students

have the knowledge of the world in their pockets (Dö-

beli, Hielscher & Hartmann, 2018, p.23).

Source: AIMDek Technologies (2018), Gould (2018), Robinson (2018) and own contributions.

Figure 2: Potential Evolution of Smart Machines.

Levelofmaturityandintegration

trans

ormative

Effectonsociety

Objective:

• Improvinglabour productivity

Features:

• Smartmachinesthatcanperform

aspecifictask(e.g. natural

languageprocessing)

• Easytouseandoperate

Limitation:

• Doesnotadjustbehaviorbased

onexperience.

AssistingSmartMachine

«supportus»

Objective:

• independentlyworking,

adaptablesmartmachine

Features:

• Adjustsbehaviorbasedon

experience

• Abilitytolearnforcertain

predefinedtasks

• PredictiveAnalytics

Limitation:

• Theabilitytolearnislimitedto

certainpredefinedtasks.

AutonomousSmartMachine

«doitwithus»

Objective:

• Deephuman‐likeunderstanding

anddecisionmakingincomplex

situations

Features:

• Smartmachineaslearning

partnerandconsultant

• Learningabilityforgeneraltasks:

smartmachineistrainedandno

longerprogrammed

• Proactiveanalytics

Limitation:

• Ethicalandlegalboundaries

CognitiveSmartMachine

«doitforus»

Autonomous

self‐learning

independent

Controlled

Static

Programmed

Paradigm Shift in Human-Machine Interaction: A New Learning Framework for Required Competencies in the Age of Artiﬁcial

Intelligence?

297

Smart machines help us to access and understand

the flood of data on the internet. Through them, new

tools are available for learning and working (Döbeli

et al., 2018, p. 17). For example, free translation tools

like DeepL (https://www.deepl.com) can be a great

help when translating a text. Although we still have

to learn languages in school, technology can influ-

ence the way we learn languages in the future. Stu-

dents can use smart machines to compensate for def-

icits and improve their strengths. Just as glasses are a

support for someone who cannot see well, smart ma-

chines could be a support for someone who cannot

write well.

Similarly, in the field of computer science, smart

machines could help us to create computer code more

easily. OpenAI’s language generator GPT-3 is able to

create computer code for web-page layouts using

prompts like “Give me a button that looks like a wa-

termelon” or “I want a blue button that says sub-

scribe” (Heaven, 2020). This still means that a com-

puter engineer has to understand the basic principles

of programming. However, the smart machine can

make his work processes much more efficient.

Living in an infosphere together with smart ma-

chines also has social and ethical implications. Smart

machines are able to imitate us better and can take on

new appearances. Computationally created virtual be-

ings like Samsung´s NEON (https://www.neon.life)

are no longer visually distinguishable from real peo-

ple. Google´s algorithm Duplex can call local busi-

nesses (e.g., to make haircut appointments) (Google

Developers, 2018). The persons called often do not

notice that they are talking to a computer. In our view,

this raises ethical concerns, especially if people are

not aware that smart machines are already capable of

such things. Smart machines, who mimic and manip-

ulate human users of social networks, raise concerns

about the manipulation of elections (Schmuck & von

Sikorski, 2020).

Smart machines are changing the information and

communication habits of the society (Döbeli et al.,

2018, p. 16). Among other things, concerns about pri-

vacy, accessibility (only for those who can afford?),

social manipulation and autonomy (Raisamo et al.,

2019, pp. 138-139) should be taken seriously. Deller-

mann et al. (2019, p. 641) point out, that the goal

should not be to maximize trust in smart machines,

but rather to “find a balance between trust and distrust

that makes it possible to leverage the potentials of AI

and at the same time avoids negative effects stem-

ming from overreliance on AI”.

In the future, we may be dependent on smart ma-

chines to a certain extent (e.g., to oversee large

amounts of data). On the other hand, smart machines

will rely on us, because they need instructions to ful-

fill their purpose. We may have to acquire new com-

petencies on how to interact and collaborate with

smart machines, that go beyond ICT- or data-literacy.

3.2 Towards an Extended Learning

Framework

With the goal to navigate through a complex and un-

certain world, the OECD (2018, p. 4) has developed

a “learning-compass” depicted in Figure 3. At the

center of the framework lies the student who faces the

challenges of the 21

century, tries to transform our

society and shape the future with the goal of individ-

ual and societal well-being. Not only students but also

teachers, school managers, parents, and communities

should be considered as learners in this context

(OECD, 2018, pp. 3-5). As can be seen from Figure

3, the student should be competent in creating new

value, reconciling tensions and dilemmas and taking

responsibility (OECD, 2018, p. 5). The OECD (2018,

p. 5) defines competencies as a combination of

knowledge, skills, attitudes and values.

First, the student should be competent in creating

new value, to provide innovative solutions at afforda-

ble costs to economic, social, and cultural dilemmas.

To do that, the student should be able to think crea-

tively, develop new ways of thinking and living as

well as invent new business models and new social

models (OECD, 2018, p. 5).

Second, the student should be competent in rec-

onciling tensions and dilemmas. In a world character-

ized by equivocality, different standpoints must be

weighed against each other. Such trade-offs could in-

volve balancing autonomy and community, innova-

tion and continuity, or efficiency and the democratic

process. To achieve that, the student must become a

system thinker, who thinks and acts in a more inte-

grated way, taking short- and long-term perspectives

into account (OECD, 2018, p. 5).

Third, the student must take responsibility for

their actions. The attempt to actively shape the future

is a process of weighing up possible risks and re-

wards. Only when we accept accountability for the

products of our work, we can shape the future by eval-

uating the future consequences of our actions (OECD,

2018, p. 6).

CSEDU 2021 - 13th International Conference on Computer Supported Education

298

Source: OECD (2018, p. 4).

Figure 3: The OECD Learning Framework 2030.

In the context of smart machines, the OECD

learning framework remains still helpful and valid.

However, as we argued in the previous sections, we

think that smart machines increasingly take over a co-

roll towards us. Through that, smart machines could

support us in different ways in navigating through a

complex and uncertain world with the goal of actively

shaping the future.

First, smart machines could perform automatable

tasks for us, as we have seen in the example of Google

Duplex, who can call local businesses (e.g., to make

haircut appointments) (Google Developers, 2018).

Baldwin and Forslid (2020, p. 9) argue that smart ma-

chines can perform many mental tasks like photo

recognition, handwriting recognition, or language

translation as well as humans. Especially repetitive,

standardizable tasks can be automated more easily

(Latham & Humberd, 2018, p. 12).

Second, smart machines could prepare infor-

mation for decision making. According to Jarrahi

(2018, pp. 581-584), smart machines have a relative

advantage in handling complex situations with large

amounts of data. While the student still has to decide

where to seek and gather data, the smart machine

could support the learner by collecting, curating, pro-

cessing and analysing data (Jarrahi, 2018, p. 583).

Ideally, the student would be able to make better de-

cisions, as the smart machine creates a more solid ba-

sis for decision making.

Third, smart machines could consult us in our pro-

cess of creating new values and reconciling tensions

and dilemmas. While the responsibility for action re-

mains in the hands of the student, the smart machine

can suggest different approaches and methods, and

weigh up possible risks and rewards as objectively as

possible. In the light of hybrid intelligence (Deller-

mann et al., 2019), the student and the smart machine

could form a team that makes them both more effi-

cient and at the same time – to a certain degree - mu-

tually dependent on each other. Figure 4 depicts this

relationship.

To a certain degree, the student is depending on

smart machines because the student needs the smart

machine to make meaningful decisions based on large

amounts of data. Further, smart machines can relieve

the student from repetitive, automatable tasks and im-

prove the decision making through consulting.

To a certain degree, a smart machine is depending

on the student, because the student is the one, who

needs the vision to create new values, reconcile ten-

sions and dilemmas along this way, and take respon-

sibility for his chosen actions. Even though the smart

machine can enhance the capabilities of the student,

the smart machine needs to be guided and checked for

potential data biases.

In our view, the paradigm shift of mutual depend-

ency means that there will be a shift in the competen-

cies required. Certain tasks and competencies are al-

ready today being outsourced to smart machines and

will probably become less important in the future

(e.g., repetitive desk research). Other competencies,

such as the capability to collaborate with smart

Paradigm Shift in Human-Machine Interaction: A New Learning Framework for Required Competencies in the Age of Artiﬁcial

Intelligence?

299

Source: OECD (2018, p. 4) and own contributions.

Figure 4: Paradigm Shift of Mutual Dependency.

machines to use them efficiently as a working tool,

will probably become more important. In the future,

people may need to acquire a “robot-proof” education

in order to stay relevant on the job market (Aoun,

2017).

This may include new knowledge in the form tech-

nical literacies (e.g., Aoun, 2017, pp. 55-58; Carretero,

Vuorikari & Punie, 2017, p. 11; OECD, 2018, p. 4) but

also new knowledge about ourselves as Aoun (2017,

pp. 58-61) argues for a new concept he calls human lit-

eracy. Wilson and Daugherty (2018, p. 11) further pro-

pose fusion skills, which are the skills that enable “to

work effectively at the human-machine interface”.

This may involve the ability to delegate tasks to smart

machines, train them, or formulating the questions in a

way, such that the smart machine can deliver the an-

swers (Wilson & Daugherty, 2018, p. 11).

To collaborate with smart machines, we might also

need new attitudes and values. In order to use smart

machines, we have to accept them (see e.g., De Graaf

& Allouch, 2013), which implies a certain level of col-

laboration readiness. Further, we need to develop an

awareness of the different situations where we can en-

counter smart machines. We need to develop a sense

for their limitations and be cautious about their poten-

tial data biases. This may also involve thinking about

explanations for the smart machine´s outcome (Wilson

& Daugherty, 2018, pp. 5-6).

We may have to rethink our attitude towards dig-

ital devices if smart machines like Amazon's Alexa

are able to listen to our conversations. “To speak

freely” could take on a different meaning in the fu-

ture. In a broader sense, as smart machines become

more similar and more compatible to us, we may have

to treat them more like equals, learn how to form ap-

propriate relationships (e.g., see Suto, 2013), and

think about augmentation strategies (Davenport &

Kirby, 2016) to position ourselves in relation to the

smart machine. Similar as the Copernican, Darwinian

or Freudian revolution, smart machines may chal-

lenge us in what it means to be human (Floridi, 2016).

4 CONCLUSIONS

Based on the OECD learning framework 2030

(OECD, 2018, p. 4), we have discussed the potential

influence of smart machines on the competency re-

quirements for the 21st century. We shed light on the

underlying factors that may lead to a paradigm shift

of mutual dependency of human and smart machine

CSEDU 2021 - 13th International Conference on Computer Supported Education

300

and outlined competencies, that may be needed in or-

der to collaborate with smart machines.

In our view, it is still an open question how to best

foster collaboration competencies with smart ma-

chines. Aoun (2017, pp. 77-110) proposes an “exper-

imental learning” approach, which integrates class-

room and real-world experiences (Aoun, 2017, p. 81).

Although we think that this is already a promising ap-

proach, more research in this area is needed.

With this paper, we want to contribute to a better

understanding of the changing human-smart machine

relationship in the age of artificial intelligence, in

order to eliminate prejudices and to lay the foundation

for better decisions on the use of smart machines. In

the light of human augmentation, young people - as

the citizens and decision makers of tomorrow –

should be equipped with the necessary knowledge,

skills, attitudes and values to recognize the opportu-

nities as well as the dangers in the use of smart ma-

chines. In this way, they could increase their ability

to actively shape the future.

REFERENCES

AIMDek Technologies (2018, August 29). Evolution of Ro-

botic Process Automation (RPA): The Path to Cognitive

RPA. Medium. https://medium.com/@AIMDekTech/

evolution-of-robotic-process-automation-the-path-to-

cognitive-rpa-c3bd52c8b865

Aoun, J. E. (2017). Robot-proof: higher education in the

age of artificial intelligence. MIT press.

Arbesman, S. (2013). The half-life of facts: Why everything

we know has an expiration date. Penguin.

Baldwin, R. (2019). Globalisation 4.0 and the future of

work. Economistas, (165), 63-75.

Baldwin, R., & Forslid, R. (2020). Globotics and develop-

ment: When manufacturing is jobless and services are

tradable (Working paper 26731). National Bureau of

Economic Research. https://doi.org/10.3386/ w26731

Brown, T.B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J.,

Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G.,

Askell, A., Agarwal, S., Herbert-Voss, A., Krueger, G.,

Henighan, T., Child, R., Ramesh, A., Ziegler, D. M.,

Wu, J., Winter, C., ... Amodei, D. (2020). Language

models are few-shot learners. arXiv preprint.

https://arxiv.org/abs/2005.14165

Carretero, S., Vuorikari, R., & Punie, Y. (2017). DigComp

2.1: The Digital Competence Framework for Citizens

with eight proficiency levels and examples of use. Pub-

lications Office of the European Union.

https://doi.org/10.2760/38842

Davenport, T. H., & Kirby, J. (2016). Only humans need

apply: Winners and losers in the age of smart machines.

HarperCollins Business.

De Graaf, M. M., & Allouch, S. B. (2013). Exploring influ-

encing variables for the acceptance of social robots.

Robotics and autonomous systems, 61(12), 1476-1486.

https://doi.org/10.1016/j.robot.2013.07.007

Dellermann, D., Ebel, P., Söllner, M., & Leimeister, J. M.

(2019). Hybrid intelligence. Business & Information

Systems Engineering, 61(5), 637-643.

https://doi.org/10.1007/s12599-019-00595-2

Döbeli Honegger, B., Hielscher, M., & Hartmann, W.

(2018). Lehrmittel in einer digitalen Welt.

Expertenbericht im Auftrag der Interkantonalen

Lehrmittelzentrale (ilz). https://www.ilz.ch/

fachberichte/

Floridi, L. (2013). The ethics of information. Oxford Uni-

versity Press. https://doi.org/10.1093/acprof:oso/

9780199641321.001.0001

Floridi, L. (2016). Hyperhistory, the Emergence of the

MASs, and the Design of Infraethics. OpenMind

BBVA. https://www.bbvaopenmind.com/en/articles/

hyperhistory-the-emergence-of-the-mass-and-the-

design-of-infraethics/

Google Developers. (2018, May 8). Keynote (Google I/O

'18) [Video]. YouTube. https://www.youtube.com/

watch?v=ogfYd705cRs&pbjreload=101

Gould, R. (2018, August 9). Robotic Process Automation

(RPA): Past, Present and Future. Kofax.

https://www.kofax.com.ru/blog/robotic-process-autom

ation-rpa-past-present-and-future

Gratton, L., & Scott, A. J. (2016). The 100-year life: Living

and working in an age of longevity. Bloomsbury Pub-

lishing.

Heaven, W. D. (2020, July 20).

OpenAI’s new language gen-

erator GPT-3 is shockingly good—and completely mind-

less. MIT Technology Review. https://www.technolo-

gyreview.com/2020/07/20/1005454/openai-machine-lea

rning-language-generator-gpt-3-nlp/

Jarrahi, M. H. (2018). Artificial intelligence and the future

of work: Human-AI symbiosis in organizational deci-

sion making. Business Horizons, 61(4), 577-586.

https://doi.org/10.1016/j.bushor.2018.03.007

Latham, S., & Humberd, B. (2018). Four ways jobs will re-

spond to automation. MIT Sloan Management Review,

60(1), 11-14.

List, A. (2019). Defining digital literacy development: An

examination of pre-service teachers’ beliefs. Comput-

ers & Education, 138, 146-158.

https://doi.org/10.1016/j.compedu.2019.03.009

Ng, W. (2012). Can we teach digital natives digital liter-

acy?. Computers & Education, 59(3), 1065-1078.

https://doi.org/10.1016/j.compedu.2012.04.016

Organisation for Economic Co-operation and Develop-

ment. (2018). The future of education and skills: Edu-

cation 2030. Directorate for Education and Skills,

OECD. http://hdl.voced.edu.au/10707/452200

Pereira, A. (2019, March 15). What are smart machines?

Career in STEM. https://careerinstem.com/what-are-

smart-machines/

Quick, M. (2019, July 22). Worklife 101: Superjobs. BBC.

https://www.bbc.com/worklife/article/20190719-su-

perjobs

Raisamo, R., Rakkolainen, I., Majaranta, P., Salminen, K.,

Rantala, J., & Farooq, A. (2019). Human augmentation:

Paradigm Shift in Human-Machine Interaction: A New Learning Framework for Required Competencies in the Age of Artiﬁcial

Intelligence?

301

Past, present and future. International Journal of Hu-

man-Computer Studies, 131, 131-143.

https://doi.org/10.1016/j.ijhcs.2019.05.008

Rios, J. A., Ling, G., Pugh, R., Becker, D., & Bacall, A.

(2020). Identifying critical 21st-century skills for work-

place success: a content analysis of job advertisements.

Educational Researcher, 49(2), 80-89.

https://doi.org/10.3102/0013189X19890600

Robinson, M. (2018, October 23). 3 reasons why your RPA

deployment will fail, and what to do about it. Medium.

https://medium.com/datadriveninvestor/3-reasons-why

-your-rpa-deployment-will-fail-and-what-to-do-about-

it-b847a30c6e94

SBFI (2019). Herausforderungen der künstlichen

Intelligenz Bericht der interdepartementalen

Arbeitsgruppe «Künstliche Intelligenz» an den

Bundesrat. Bern. https://www.sbfi.admin.ch/sbfi/de/

home/bfi-politik/bfi-2021-2024/transversale-themen/di

gitalisierung-bfi/kuenstliche-intelligenz.html

Schmuck, D., & von Sikorski, C. (2020). Perceived threats

from social bots: The media's role in supporting liter-

acy. Computers in Human Behavior, 113, 106507.

https://doi.org/10.1016/j.chb.2020.106507

Suto, H. (2013). Robot Literacy An Approach For Sharing

Society With Intelligent Robots. International Journal

of Cyber Society and Education, 6(2), 139-144.

http://dx.doi.org/10.7903/ijcse.1057

Wesche, J. S., & Sonderegger, A. (2019). When computers

take the lead: the automation of leadership. Computers

in Human Behavior, 101, 197-209.

https://doi.org/10.1016/j.chb.2019.07.027

Wike, R., & Stokes, B. (2018, September 13). In Advanced

and Emerging Economies Alike, Worries About Job Au-

tomation. Pew Research Center. https://www.pewre-

search.org/global/2018/09/13/in-ad

vanced-and-emerging-economies-alike-worries-about-

job-automation/

Wilson, H. J., & Daugherty, P. R. (2018). Collaborative in-

telligence: humans and AI are joining forces. Harvard

Business Review, 96(4), 114-123.

Zornoza, J. (2020, April 15). Explainable Artificial Intelli-

gence: Discover one of the biggest trends in Machine

Learning and AI. Towards data science. https://to-

wardsdatascience.com/explainable-artificial-intelli-

gence-14944563cc79

CSEDU 2021 - 13th International Conference on Computer Supported Education

302