Artificial Intelligence in Organizations

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1393
Artificial Intelligence in Organizations
Namdev Vishnu Warang
M.sc in Information technology, Keraleeya Samajam’s Model College, Dombivli Maharashtra, India
---------------------------------------------------------------------***----------------------------------------------------------------------
Abstract - Artificial Intelligence (AI) technologies provide
each novel distinctive opportunities and cause new significant
challenges to organizations that set them aside from different
styles of digital technologies. This article discusses thedistinct
effects of AI technologies in organizations, the tensions they
raise and the opportunities they gift for info systems (IS)
analysis. we tend to explore these opportunities term of 4
capabilities automation, engagement, insight/decision
creating and innovation. we tend to discuss the differentiated
effects that AI brings concerning and also the implications for
IS future analysis.
Key Words: Artificial Intelligence,researchagenda,Machine
learning, algorithms, robotic automation, conversational
agents, robotics, AI capabilities
1.INTRODUCTION
The artificial intelligence (AI) pioneers of the Fifties
envisioned building machines that would sense, reason, and
suppose like folks. whereas such a vision remains in the
realms of fantasy, fashionable advances in computing and
therefore the omnipresent handiness of huge datasets have
allowed organizations to implement AI technologies that
transcend automating and informating. Recently developed
AI agents square measure capable of “learning,” resolution
issues, recognizing and displaying emotions, and making
outcomes in increasingly numerous domains, from
developing new products to autonomously managing
business processes and providechains(Daugherty&Wilson,
2018). for instance, machine learning algorithms detect
suspicious money transactions and advocate selections to
manage fraud (Davenport,2018). Sensible bots and vehicles
square measure autonomously delivering food and drugs.
Robots and machines serve as reliable companions,
responding to human emotions, respondent queries, and
giving help in diverse settings (e.g., isolated elderly).
AI technologies provide each novel distinctiveopportunities
and cause new and important challenges to organizations in
ways in which take issue from different digital technologies.
First, AI technologies take issue in their capacity to
constrain, complement, and/or substitute for humans at
work once they're deployed in associate organization
(Murray et al., 2020). This shifts the locus of action, choice,
control, and power far from the exclusivedomainofhumans,
requiring the development of associate understanding of
however humans and AI technologies act in new ways in
which to supply a stabilizing force, a coevolution of labor, or
the emergence of novel styles of work and organizing
Second, AI technologies essentially challenge our long-held
beliefs dividing the realms of human ability and machine
capabilities (Schuetz & Venkatesh 2020). Recent AI
technologies square measure capable of playing various
human feats, like perception, sensing and recognizing
emotions, spoken language, and even creativity. Such new
capabilities permit AI to enter domains that have up to now
remained exclusive to humans (e.g., recursive management,
new product development, and emotions recognition).
Although how machines ought to behave or suppose
continues to be controversial, recent advances in AI
capability invoke many tensions that transcend human-
machineinteractions ornew human-machineconfigurations.
Third, AI technologies exhibitincreasinglevelsof complexity
that usually result in several surprising twin outcomes
(Benbya et al., 2020b). While AI technologiesprovideseveral
positive advantages to organizations, their introduction
typically creates significant fortuitous (or intended)
consequences for individuals and organizations. Since the
impact of AI implementation varies greatly among
stakeholders, decisions to decouple stakeholders from the
method of designing, implementing, and mistreatment AI
systems typically lead to the last word failure of systems
(Wright & Schultz 2018). To account for such complexness
given the wide spectrum of stakeholders concerned
warrants a multi-stakeholder perspective(Clarke&Davison
2020). The distinct effects ofAItechnologiesinorganizations
present opportunities for data systems (IS) research. We
tend to explore these opportunities in termsof four business
capabilities: automation, engagement,insight/decision-
making, and innovation. We discuss theimplicationsforISof
the differentiated effects engendered by AI. Before doing
thus, we tend to in brief discuss the evolution of AI
technologies.
2. Developments in Artificial Intelligence
Although the origin of computing (AI) dates to the
seventeenth century, the AI field emerged within the 70s
from research into developing machines able to perform
humanlike psychological feature tasks (e.g., thinking,
learning, and conversing), spanning contributions from
numerous fields like biology, linguistics, psychology,
cognitive sciences, neurobiology, arithmetic, philosophy,
engineering, and computing. Early efforts in computing

geared toward building machines capable of simulating
human intelligence. Despite such tries and guaranteesofthe
sensible usefulness of AI, it for the most part didn't deliver
and moon-faced several obstacles, significantly throughout
the Nineteen Sixties and1970s, the largest of that was the
shortage of computational power to try to to something
substantial. During the Eighties and Nineteen Nineties,
skilled systems emerged as sensible applications supported
earlier analysis in AI. And, within the early 2000s, machine
learning and neural networksbegantoflourishascompanies
integrated statistics and likelihood into numerous business
applications. Over successive decade, digital systems,
sensors, and therefore the net proliferated, providing all
kinds of knowledge for machine learning consultants to use
to train adjustive systems. though the expansion of AI and
machine learning has been intermittent, the present
unprecedented computing capability and growing volumes
of knowledge have semiconductordiodetotheemergence of
contemporaryAItechnologies.Informationsystemsstudents
have a protracted history of conducting analysis on
computing. IS as a discipline emerged once computers
enabled the automation of business processes andtherefore
the digital capture of business transactions. IS analysisonAI
has been conducted since the Nineteen Seventies, with early
developments in call support systems (Alter, 1978), skilled
systems and knowledge-based systems (Meyer & Curley,
1991), and, later, recommendation agents (Xiao & Benbasat,
2007). Such systems, however, weren't capable of
mechanically learning and up their strategies and were
dependent on human programmers to regulate them. In
distinction, more contemporary AI technologies area unit
designed not solely to help managers with repetitivechoices
and sophisticated unstructured issues however are capable
of learning, adjusting their behaviors, and creating
autonomous complex choices. Such technologies embrace
machine learning (and its deep learning and reinforcement
learning subclasses), linguistic communication process,
robots, numerous automation technologies (including
robotic method automation), and rule-basedskilledsystems
(still in broad use though not thought-about a state-of-the-
art technology). Table A1 within the Appendix provides
transient definitions, domain of applications, and
classifications of various AI technologies in organizations.
3. Research Opportunities Enables Organizations and
Business Capabilities
AI technologies area unit progressively overlapping and
becoming embedded inside completely different structure
applications (Davenport, 2018). instead of narrowing our
specialize in one distinct technology(e.g., machine learning),
we tend to examine analysis opportunities according to the
subsequent completely different business capabilities:
Journal of the Association for data Systems 3
• Automation of structured (or semistructured)work
processes, typically via artificial intelligence,robotic process
automation, machine learning, and rule based systems.
• Engagement with customers and workers, using linguistic
communicationprocesschatbots,intelligentagents,machine
learning, and computer vision.
• Decision-making through intensive analysis structured
information, most frequently victimizationmachinelearning
algorithms and neural networks.
• Creation of novel outcomes by combining machine
learning, neural networks, and computer vision.
Such AI-enabled capabilities area unit on-going, dynamic,
overlapping processes between completely different
sociotechnical and data-related entities and therefore the
tensions that emerge from their manifold interactions
though these andalternative capabilitieslikeinnovationarea
unit typically combined or presented at the same time, for
the sake of simplicity, we will discuss every of the
capabilities and theirassociatedtensionsseparatelythen gift
related analysis queries.
4. AI-Enabled Automation
AI-enabled automation revolves round the use of
technologies to support structured and semistructured
tasks. These tasks ar usually repetitive, labor intensive, and
embody physical moreover as psychological feature tasks.
Performing physical tasks is that the ancient domain of
robots in settings like works automation. AI enabledrobots
ar equipped with the power to sense their surroundings,
comprehend, act, and learn. This helps robots perform
several tasks by with success navigating their surroundings,
characteristic objects around them, and helping humans
with numerous tasks such as autonomous deliveries and
robot-assisted surgeries (Benbya et al. 2020). psychological
feature automation consists in mistreatment technologies
like robotic automation or machine learning technologies.
Robotic process automation (RPA) usually automates
routine administrative tasks (e.g., knowledge entry work)
(Lacity & Willcocks, 2016), whereas machine learning is
employed to analyze and establish anomalies in massive
datasets and increase thespeed,roughness,andproductivity
of modeling. Developing such technologies in organizations
to modify automation capabilities invokes many tensions
regarding however work is performed. Below, we have a
tendency to discuss a number of them.
5. Research Opportunities
As completely different AI technologies area unitintroduced
to substitute for varied tasks, opportunitiestodeal withhow
such technologies become integrated among the
organization area unit incessantly arising. IS researchers

focusing on adoption may specialize in the characteristics or
options of AI technologies that increase acceptance and use.
as an example, the visibility of the work carried out by
physical robots might trigger staff and managers to
additional simplyacknowledgethe worthofautomatingsuch
physical tasks. Pachidi et al (2020)found that once a robotic
method automation tool runs in the background, it should
become tougher for employees to forgoing of psychological
feature tasks within which they have invested with data and
experience. Task automation implies increasing interaction
of humans with machines. this kind of interaction might
differ if one focuses on physical robots versus robotic
process automation tools. Physical robotsarea unitseenand
felt by staff and their physical activity causes visiblechanges
within the physical atmosphere of the workplace. IS
researchers that specialize in human-machine interaction
may study very well however staff move with physical
robots, and the way they alter their routines in order to
accommodate robots’ movements within the workspace. In
distinction, robotic method automation tools might not be
visible to staff and their algorithms area unit doubtless tobe
black-boxed to them. Researchers may investigate the
challenges that workers face as they move with automation
tools that automate varied tasks or the outputs created by
those tools. probably, staff might return to develop various
workarounds so as to beat difficulties. As AI-enabled
automation technologies become further enforced, we tend
to area unit doubtless to visualize changes in organizational
communication. as an example, the use of robotic method
automation tools can doubtless alter information flows
within the organization, resultingintheintegrationofrecent
roles targeted on configuring automation tools and act
effectively with other stakeholders. AI-enabled automation
technologies also can trigger vital changes in how
coordination is achieved among human consultants. For
example, Sergeeva et al. (2020) illustrate the redistribution
of tasks resultant from the introduction of robots in medical
operations. The coordinating adaptations they examined
eventually light-emitting diode to the reconfiguration of
roles, growth of activity knowledge, and shifts in activity
boundaries and status arrangements. However, additional
still must be learned regarding this. as an example, however
do different less tangible styles of automation technologies
like algorithms have an effect oncoordinationamonghuman
experts? How can coordination amendment as human
consultants begincollaboratingwithautomationtools?What
area unit the characteristics of automation tools which will
form coordinative adaptations?
6. AI-Enabled Engagement
AI-enabled engagement refers to the final capability of
computers to grasp, respond, engage, and converse with
humans victimization natural human language. Although
such engagement includes each voice- and text-based
technologies, the technologies used dissent largely
supported their capability,domain,andlevel ofembodiment.
easy AI engagement technologies are mainly accustomed
handle repetitive shopper queries whereas smarter
technologies, enabled by machine learning and natural
language process, have the potential to undertake a lot of
advanced tasks that involve bigger interaction, voice
communication, reasoning, prediction, accuracy, and
emotional show. Such technologies have been employed in
many alternative fields, including finance, commerce,
marketing, retail, and aid. Although the technologies behind
AI-enabled engagement are endlesslybeneathdevelopment,
they currently don't have full human-level languageabilities,
generally leading to misunderstanding and user
discontentment.
7. AI-Enabled Insights and Decisions
AI-enabled insights revolve round the use of machine
learning (ML) algorithms—a set of unambiguous
instructions that a mechanical laptop will execute. Some
millilitre algorithms are often trained on structured
knowledge and ar specific to slender task domains, like
speech recognition and image classification. Different
algorithms, particularly deep learning neural networks, can
learn from giant volumes of tagged knowledge, enhance
themselves by learning, and attain a range of tasks like
classification, prediction, and recognition. For instance,
neural networks will analyse parameters of bank shoppers
like age, solvency, and credit history to determine whether
or not to approve a loan request. Such networks can even
use face recognition to permit solely approved individuals
into a building or predict outcomes like theincreaseorfall of
a stock supported past patterns and current knowledge.
8. AI-Enabled Innovation
Beyond the 3 business capabilities—AI-enabledautomation,
AI-enabled engagement, and AI-enabled insight—there ar
alternative business capabilities like innovation. Machine
learning and deep learning neural networks will alter or
enhance innovation processes and outcomes.AIdata-driven
insights, models, and visualizations will facilitate theartistic
interpretation of information and support decision-making
within the innovation method (Wu etal.,2020).Finally,deep
learning has the potential to shorten the time needed to
bring new merchandise to markets. As a result, many
pharmaceutical firms and biotech start-ups have invested
with in AI to spot and validate potential drug candidates to
accelerate the general drug discovery method (Fleming
2018). Though AI technologies might not nevertheless be
able to severally develop entire solutions, they'll purpose
human managers toward the foremost promising avenues
for innovation. yet, the employment of AI for innovation
triggers many tensions.

9. Conclusion
Many lessons are usually learned from the past successes
and failures of AI. Rational and harmonic interactions are
needed between application-specific comes and visionary
analysis ideas to sustain the progress of AI. a transparent
strategy is needed to consider the associatedmoral andlegal
challenges to confirm that the society as a full can like the
evolution of AI and its potential adverse effects are eased
from early on. Beside the unprecedented enthusiasm of AI,
there are also fears regarding the impact of technology on
our society. Such fears mustn't hinder the progress of AI
however encourage the development of a scientific
framework on that future AI can flourish. Mostvital ofall,it's
necessary to apart science fiction from sensible reality.With
sustained funding and accountable investment, AI is getting
ready to rework the long run of our society, our economy,
and our life.
References
1) Adam, M., Wessel, M., & Benlian, A. (2020).AI-based
chatbots in customer service and their effects on
user compliance. Electronic Markets, 17, 1-19.
2) Agrawal, A., Gans, J., & Goldfarb, A. (2018).
Prediction machines: The simple economics of
artificial intelligence.
3) Harvard Business Review.
4) Alter, S. (1978). Development Patterns for Decision
Support Systems Steven Alter, MIS Quarterly, 2(3),
33-42.
5) Amershi, S., Cakmak, M., Knox, W. B., & Kulesza, T.
(2014). Power to the people: The role of humans in
interactive machine learning. AI Magazine, 35(4),
105–120.
6) Anthony, C. (2018). To question or accept? How
status differences influence responses to new
epistemic technologies in knowledge work.
Academy of Management Review, 43(4), 661-679.
7) Bader, V., & Kaiser, S. (2019). Algorithmic decision-
making? The user interface and its role for human
involvement in decisions supported by artificial
intelligence. Organization, 26(5), 655-672.
8) Balasubramanian, N., Ye Y, & Xu M. (in press).
Substituting human decision-making with machine
learning: implications for organizational learning.
Academy of Management Review.
9) Barrett, M., & Oborn, E., Orlikowski, W. & Yates, J.
(2012). ‘ReconfiguringBoundary Relations:Robotic
Innovations in Pharmacy Work’ Organization
Science. 23(5), 1448-1466.
10) Beane, M. (2019).Shadowlearning:Buildingrobotic
surgical skill when approved means fail.
Administrative Science
11) Quarterly, 64(1), 87-123.
12) Benbya, H., & McKelvey B. (2006) Using
coevolutionary and complexity theories to improve
IS alignment: A multilevel approach. Journal of
Information Technology, 21(4), 284-298.
13) Benbya, H., Davenport, T.H., & Pachidi, S. (2020b).
Artificial intelligence inorganizations:Currentstate
and future opportunities. MIS Quarterly Executive,
19(4).9-21.
AUTHOR:
Namdev Vishnu Warang.
B.sc.{computer science}
Pursuing M.sc {Information Technology}

Artificial Intelligence in Organizations

More Related Content

Similar to Artificial Intelligence in Organizations (20)

More from IRJET Journal (20)

Recently uploaded (20)

Artificial Intelligence in Organizations