=Paper=
{{Paper
|id=Vol-3723/paper21
|storemode=property
|title=Application of the analytical hierarchy process to evaluate hype AI technologies
|pdfUrl=https://ceur-ws.org/Vol-3723/paper21.pdf
|volume=Vol-3723
|authors=Dmytro Dosyn,Oleh Karyy,Ihor Novakivskyi,Maryana Gvozd,Yaroslav Kis,Nataliia Kara
|dblpUrl=https://dblp.org/rec/conf/modast/DosynKNGKK24
}}
==Application of the analytical hierarchy process to evaluate hype AI technologies==
https://ceur-ws.org/Vol-3723/paper21.pdf
Application of the analytical hierarchy process to
evaluate hype AI technologies
Dmytro Dosyn1,†, Oleh Karyy1,†, Ihor Novakivskyi1,†, Maryana Gvozd1,†, Yaroslav Kis1,†
and Nataliia Kara1,∗,† 1
1 Lviv Polytechnic National University, Stepana Bandery St. 12, 79013 Lviv, Ukraine
Abstract
The current topic of research on the implementation of hype technologies in the field of Artificial
Intelligence (AI) is considered. An overview of the prospects for the intensive development of the AI
field is presented. The article shows the growing attention to the introduction of AI technologies by
business structures that invest more and more money in this industry. The essence of the maturity
curve of Hype Cycle technologies developed by the Gartner consulting company is revealed. A
comparative analysis of Hype Cycles for Artificial Intelligence for 2016 and from 2019 to 2023 is
carried out. The application of the analytical hierarchy process for the selection of promising hype AI
technologies in the activities of domestic companies is substantiated.
Keywords
Artificial Intelligence, innovation, company, business processes, development prospects, hype
technologies, fuzzy systems, analytical hierarchy process.
1. Analysis of prospects for the development of AI technologies
Globally, there is an explosive interest in the growing impact of AI. The emergence of new AI
technologies affects individual consumers and companies as well as countries in general. Today,
AI technologies have become a driving force of innovation, and therefore companies are forced to
invest and implement new AI technologies to remain competitive. The development and active
use of AI technologies allow organizations to realize potential opportunities that can become
objects for improvement and implementation in their activities and ensure their success in a
dynamic business environment. The relevance of the problem is confirmed by the fact that the
world is witnessing rapid growth in AI spending, the transformation of IT services, and the
convergence of various AI technologies. According to the data of well-known analytical agencies,
the average annual growth of the global AI market in 2023 is within 20-38% (Table 1). Other
studies demonstrate that from 2018 to 2025, the market of AI technologies is expected to grow
from $21.46 billion to $190.61 billion, and the cumulative annual growth rate will make up 36.62
percent [1].
The consulting company UBS predicts that the AI industry will have grown to a $225 billion
market by 2027 and expenditures on the AI infrastructure are expected to grow from $25.8
billion in 2022 to $195 billion in 2027 [3].
The IDC Company published a forecast on the European AI market. According to the
estimates, spending on AI in Europe has reached $34.2 billion according to the results of 2023,
which has made up approximately 20.6% of the global volume. The CAGR Indicator (Compound
MoDaST-2024: 6th International Workshop on Modern Data Science Technologies, May, 31 - June, 1, 2024, Lviv-Shatsk,
Ukraine
∗ Corresponding author.
† These authors contributed equally.
dmytro.h.dosyn@lpnu.ua (D. Dosyn); oleh.i.karyi@lpnu.ua (O. Karyy); ihor.i.novakivskyi@lpnu.ua (I. Novakivskyi);
mariana.y.hvozd@lpnu.ua (M. Gvozd), yaroslav.p.kis@lpnu.ua (Y. Kis); nataliia.i.kara@lpnu.ua (N. Kara)
0000-0003-4040-4467 (D. Dosyn); 0000-0002-1305-3043 (O. Karyy); 0000-0003-0841-3603 (I. Novakivskyi);
0000-0001-7842-694X (M. Gvozd); 0000-0003-3421-2725 (Y. Kis); 0000-0001-7000-2931 (N. Kara)
© 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
Annual Growth Rate) from 2022 to 2027 in the European market is forecast at 29.6% against
26.9% globally. As a result, in 2027 AI expenses in Europe will exceed $96.1 billion per year [4].
Table 1
Average annual growth rate of the global AI market, forecasts for 2023
Consulting company Average annual growth
Precedence Research 38%
Statista 38%
IDC 27%
Fortune Business Insights 20%
Source: News Analytics, 2023 [2]
Bill Gates [5] said a personal digital agent could kill Google Search and Amazon. The first
company to develop it will have a leg up on competitors. There is a 50-50 chance that this future
AI winner will be either a startup or a tech giant. “Whoever wins the personal agent, that’s the
big thing, because you will never go to a search site again, you will never go to a productivity
site; you’ll never go to Amazon again”, he said during a Goldman Sachs and SV Angel event.
AI encompasses a wide range of methods and disciplines, including vision, perception,
language and dialogue, decision-making and planning, problem-solving, robotics, and other
areas of application where self-learning is possible. AI is optimally considered as a set of
technologies and methods used to supplement traditional human qualities, such as intelligence,
analytical, and other abilities. The results of the research made it possible to distinguish the
following main types of AI:
• Artificial Narrow Intelligence (ANI) is aimed at very good performance of a certain task (for
example, based on Natural Language Processing (NLP): Google Assistant, Google Translate, Siri,
Cortana).
• Artificial General Intelligence (AGI) represents the behavior of a human specialist. These
technologies are now intensively developing, but the real achievement of the level of a high
professional will not happen relatively soon.
• Artificial Super Intelligence (ASI) is a superintelligence that can be applied to solve any
problem. Currently, this is a distant prospect for at least several decades (according to experts,
not earlier than 2050).
Achieving the level of AGI in practice is the first ambitious task of today, but it also requires
huge resources. Therefore, regardless of how close existing AI solutions have reached the
embodiment of the highest levels of AI, they fulfill the very cutting-edge of engineering thought.
Modern AI requires high-performance computing, powerful IT infrastructure, advanced
scientific developments, and appropriate personnel. And this de facto reflects the degree of
technological maturity of companies, industries, and states.
In many sectors, AI offers innovative services; in particular, it is used for statistical and big
data analysis. AI is also used to analyze trends and identify crisis threats, optimize the business
model, and develop regulatory approaches.
IDC’s Future Scape predictions provide valuable insights into the future of the IT industry
and the pivotal role of artificial intelligence. IDC predicts significant changes in the global
business ecosystem caused by the development of the IT industry. In the world, the network
information infrastructure will be significantly rebuilt, satellite-based Internet connectivity will
be developed, Big Data processing will be intensified, and the use of AI will also expand. The
convergence of new IT capabilities will primarily contribute to the transformation of
management tools [7]. Because of this, new requirements for updating personnel’s IT skills will
be formed [8]. IDC predicts that by 2025, the world’s 2,000 largest companies (G2000) will have
allocated more than 40% of IT spending to AI [9].
By 2025, 75% of customer interactions will consist of AI contacts. Already now, analytical
robots calculate a business development plan for the next 10 years in a matter of seconds and
chatbots solve typical customer problems at any time of the day [1].
56%
60% 51%
47% 46%
50% 40%
35% 33%
40% 30% 30%
26% 24%
30%
20%
10%
0%
Figure 1: Top Ways Business Owners Use Artificial Intelligence [6].
Source: Forbes Advisor Embed
In Ukraine, the following most significant areas of the application of AI technologies can be
distinguished [10-18, 19, 20]:
1. Logistics:
• vehicle autopilot;
• tracking the movement of goods;
• formation of cargo flow traffic;
• management and optimization of stocks.
2. Finances:
• diagnosis of the company’s local financial system;
• automatic management of portfolios;
• analysis of financial risks;
• prevention of fraud and other cyber threats.
3. Management of production processes:
• diagnosis and support of technological processes;
• automation of the assembly and packaging of a product,
• warning about equipment breakdowns in production.
• automation of reporting;
• guaranteeing safety at work.
4. Organizational management system:
• support for making management decisions;
• speech recognition and command execution;
• processing and creation of graphic images and video objects;
• generation of texts, programs, images, and video plots on a given topic;
• automation of work with personnel (creating vacancies, selection of resumes, etc.);
• prototyping and personalization of users’ interfaces
5. Marketing:
• selection and recommendation of content on a certain topic (social networks, online
stores, etc.);
• analysis of big data, their segmentation, and search optimization;
• testing and optimization of advertising campaigns;
• demand forecasting;
• smart technologies for communication with customers.
• planning publication in social networks;
• personalization of customer service;
• development of animation, video, and visual solutions.
6. Medicine [19]:
• diagnosis and prognosis of diseases;
• personalized treatment;
• automatic analysis of medical images, such as radiography, CT, MRI, etc., to detect
anomalies and pathologies;
• patient monitoring and health forecasting;
• medical documentation and information processing
AI changes not only the IT industry but also takes business development to a
qualitatively new higher level. In particular, experts claim that after 2014, the most significant
developments in the IT field are carried out by commercial structures, not academic institutions
(non-commercial and scientific organizations). This is because the creation of modern AI
systems requires huge volumes of data, as well as large computing and financial resources.
In general, AI is not a separate factor in solving specific tasks, but through convergence plays
the role of a catalyst and amplifier of existing technologies, transferring them to a qualitatively
higher level of utility for the consumer. The practical aspects of using AI technologies are
expanding rapidly. This is confirmed by the research of scientists [21] who offer a
comprehensive review of approaches to the analysis of failure in industrial systems using
artificial intelligence with sufficient or insufficient data and unbalanced problems. The scientists
have presented AI algorithms and classified the scenarios of industrial system applications into
homogeneous and heterogeneous ones based on data. The authors also summarize the solved
problems, challenges, and promising directions. This can become a new direction for the
practical application of AI technologies.
2. Summary analysis of Hype Cycles for Artificial Intelligence for 2016,
2019-2023.
To make the most of new possibilities of hype AI technologies, it is extremely important to
see the ways of their mass implementation and cases of potential use. Individual AI technologies
are at an early stage and there is great uncertainty regarding their development. Their
introduction to the market can last up to 10 years. It is clear that focusing on hype AI
technologies is associated with great risks during deployment, but potentially has more benefits
for early adopters.
In 1995, the consulting company Gartner proposed the term Hype Cycle for planning, which
became the name for the maturity curve of technologies. Gartner presents research materials in
annual analytical reports [22-27], which show about 30 new technologies that can greatly affect
consumers, businesses, society, and people over the next 5-10 years.
The consulting company Gartner considers Hype technologies to be revolutionary
technologies whose viability and competitive advantage have not been proven. The company’s
analysts are involved in the research, sifting through thousands of unique technologies that, in
their opinion, will manifest themselves within 5-10 years and will have a strong impact on
society and business. This schedule is divided into five stages, which are shown in Table 2.
Table 2
Hype Cycles content
№ Stage Content
A technological breakthrough in the development of hype
Innovation technologies is advertised. The first tests of its technology
I
Trigger functionality are conducted. Sometimes usable products do not exist
and commercial viability is not proven.
Peak of The first single successful implementations of hype technologies
II Inflated are widely advertised. Mass applications of these technologies have
Expectations not yet confirmed the expectations of consumers who are waiting.
III Trough of Interest in the use of new hype technologies declines, as mass
Disillusionment experiments do not confirm the expected results. Investments in
development continue if the manufacturers of hype technology
improve it to the level that satisfies the first at least partially contented
users.
The possibilities are clarified and the scope of the application of
new hype technologies is expanded. Hype technology providers are
Slope of
IV developing second and third-generation products. More and more
Enlightenment
conservative companies that have taken a wait-and-see position
introduce new hype technologies.
The market for the consumption of new hype technologies
stabilizes. The criteria for assessing the viability of suppliers are more
Plateau of
V defined. The wide application and relevance of hype technologies on
Productivity
the market pay off. The evolution of hype technologies and their
spread into new market segments is possible.
Source: [28]
Thirty points are drawn on this graph, which is tied to specific technologies and
indicates the expected time of their release to mass sales. In particular, the Hype Cycle chart:
• shows the horizons of technological prospects and makes adequate decisions on the use
or non-use of novelties;
• reflects the relative location of various adjacent or competing technologies, the level of
their viability, and the speed of adaptation;
• provides insight into how technology will evolve, providing a reliable source of
information for decision-makers;
• helps determine the optimal time for investments and reduces the risks of unsuccessful
investments.
The Hype Cycle methodology helps professionals evaluate the potential opportunities of
new technologies for business.
The growing attention to the development of AI is confirmed by Gartner’s annual
presentation on the hype analysis of AI technologies, starting in 2019.
The summarized results of Gartner’s analysis of Hype Cycles for Artificial Intelligence from
2016 to 2023 by development stages are shown in Table 3 [22-27].
Table 3
Hype Cycles for Artificial Intelligence
2016 2019 2020 2021 2022 2023
Stage I. Innovation Trigger
Automatic ③
Smart Dust ④ AGI ④ AGI ④ AGI ④ AGI ④
Systems
AI Physics- Physics- First- ②
4D Printing ④ ③ Small Data ③ ③ ②
Marketplaces Informed AI Informed AI Principles AI
Reinforceme Multiagent ③
G-PMI ④ ③ Composite AI ② AI TRiSM ③ Causal AI ③
nt Learning Systems
AI Data-Centric Neuro- ④
802.11ax ③ ③ Generative AI ② Composite AI ② ②
Governance AI Symbolic AI
Context Augmented AI Model AI ②
③ ② ③ ③ ③ Causal AI
Brokering Intelligence Marketplaces Compression Engineerings
Responsible Small and Decision ③
NC ④ NC ③ ③ ③ ② AI Simulation
AI Wide Data Intelligence
Decision Things as ③
dbrPaaS ③ ③ ④ AIO&AP ② Composite AI ② AI Engineering
Intelligence Customers
Quantum AI Cloud Machine Data-Centric ②
④ ③ NC ③ ④ AI TRiSM ②
Computing Services Customers AI
Human ④ DL&A ② Augmented ② ModelOps ③ Operational ③ Composite AI ②
Augmentation Services Intelligence AI Systems
Personal Knowledge AI Responsible Operational AI ③
③ ③ ② ③ NCMOps ③
Analytics Graphs Governance AI Systems
Smart AI-Related Multiexperie ②
③ ② AI D&TK ② ③ AI TRiSM
Workspace C&SI Services nce
Volumetric AI Developer AI Decision ②
④ ② ②
Displays Toolkits Governance Intelligence
Explainable Generative ④
CUI ③ ③ ② AGI
AI AI
Human-
B-CI ④ Edge AI ② ②
Centered AI
Smart
VPA ③ ③ NC ③
Robots
Smart Data Synthetic
③ ②
Discovery Data
Stage II. Peak of Inflated Expectations
Decision Decision Prompt ②
Smart Robots ③ DNN ASICs ② ② ② Generative AI ②
Intelligence Intelligence Engineering
Transformer Responsible ③
Blockchain ③ AI PaaS ③ Smart Robots ③ ③ ③ NC
s AI
Connected Quantum DL&A Foundation Responsible ③
③ ④ ② Smart Robots ③ ③
Home Computing Services Models AI
Cognitive ③
Intelligent Knowledge
Expert ③ ② DNN ASICs ② ③ Smart Robots ③ Smart Robots
Applications Graphs
Advisors
Machine Intelligent Synthetic Foundation ②
② Digital Ethics ③ ② Edge AI ② ②
Learning Applications Data Models
Knowledge ③
SDS ③ AutoML ② ③ AIM&TK ② Edge AI ② Generative AI
Graphs
Autonomous Knowledge ②
② Chatbots ② Digital Ethics ③ DNN ASICs ② ③ Synthetic Data
Vehicles Graphs
Nanotube
③ CUI ③ Edge AI ② Digital Ethics ③
Electronics
(Deep AI Cloud AI Cloud
SDA (SDx) ② ② ② ②
Learning Services Services
Graph Deep
③ Deep Learning ② ②
Analytics Learning
Machine
② NLP ③
Learning
Stage III. Trough of Disillusionment
Machine DL&A ③
NLQA ② NLP ③ ② ② NLP ③ ModelOps
Learning Services
FPGA ①
ET&OM ④ VPA-EWS ② ② NLP ③ Digital Ethics ② EdgeAI
Accelerators
Augmented Machine Knowledge ②
③ RPAS ① Chatbots ② ② AIM&TK ②
Reality Learning Graphs
FPGA Computer Intelligent AI Cloud ②
② ② ② ② AIM&TK
Accelerators Vision Applications Services
Virtual Autonomous Deep Autonomous ③
② ④ Chatbots ① ②
Assistants Vehicles Learning Vehicles
Computer Cognitive Autonomous Autonomous
② ⦻ ④ ④
Vision Computing Vehicles Vehicles
Insight Computer
② ②
Engines Vision
Cognitive
③
Computing
Autonomous ④
Vehicles
Stage IV. Slope of Enlightenment
Insight Sematic Intelligent Intelligent ②
Virtual Reality ③ ② ② ②
Engines Search Applications Applications
Cloud AI ②
DL&A ②
Services
Computer ①
① DL&A
Vision
Computer ①
Vision
Stage V. Plateau of Productivity
Speech GPU
① ①
Recognition Accelerators
GPU
①
Accelerators
AGI - Artificial General Intelligence;
AI D&TK - AI Developer and Teaching Kits;
AIM&TK - AI Maker and Teaching Kits;
AIO&AP - AI Orchestration and Automation Platform;
B-CIe - Brain-Computer Interface;
CUI - Conversational User Interfaces;
dbrPaaS - Data Broker PaaS;
DL&A - Data Labeling and Annotation;
DNN ASICs - Deep Neural Netwrok ASICs;
ET&O - Enterprise Tewonomy and Ontology Management;
G-PMI - General-Purpose Machine Intelligence;
NC - Neuromorphic Computing;
NCMOps - Neuromorphic Computing ModelOps ;
NLP - Natural Language Processing;
NLQA - Natural-Language Question Answering;
RPAS - Robotic Process Automation Software;
SDA (SDx) - Software-Defined Anything (SDx);
SDS - Software-Defined Security;
VPA - Virtual Personal Assistants;
VPA-EWS - VPA-Enabled Wireless Speakers;
Source: developed by the authors on the base of [22-27].
Plateau will be reached: ① - less than 2 years; ② - 2 to 5 years; ③ - 5 to 10 years; ④ - more
than 10 years; ⦻ - obsolete before plateau.
However, despite its great advantages, the Hype Cycle chart, like any analytical tool, cannot
be the final criterion of truth, as can be seen from the analysis in Table 3. For example, let the
authors consider Cognitive Computing, which simulates human mental processes in computer
systems, and by 2019, they occupied a fifth of the AI market. Currently, developments in this
area have moved into a set of narrower problems for solving narrow business problems (for
example, fraud detection and risk management, attracting customers with the help of
personalized marketing, monitoring the condition of equipment or product quality, adjusting
technological processes, diagnosing diseases, etc.). Currently, the cognitive computing market
can be divided into four segments: natural language processing; information search; machine
learning; and automated thinking. The natural language processing (NLP) segment captured the
largest market share in 2022 due to the massive adoption of chatbots and voice assistant
devices.
In general, the number of long-term projects of hype AI technologies is reduced, the
emphasis is shifted to more specific goals and no positions are taken on the Stage V Plateau of
Productivity of Hype Cycles for Artificial Intelligence (except for 2019 − GPU Accelerators).
3. Application of the analytical hierarchy process for the selection of
promising hype AI technologies for the organization
In general, it can be stated that the elements of AI are increasingly penetrating the activities
of all organizations since Artificial Intelligence (AI) is an evolving set of technologies used for
solving a wide range of applied issues [29]. Their scope of application is increasing every year.
Therefore, companies face the question of choosing the AI technologies that are most correctly
integrated into their business processes. Such a choice lies in the following two areas: 1) the
specifics of the organization’s activity; 2) the dynamically growing capabilities of AI
technologies. The choice of the company’s policy in both areas can be based solely on subjective
judgments. Managers and interested stakeholders understand the company’s tasks best. The
most complete assessment of the forecast of the influence of AI development technologies can
be estimated exclusively from the data of the analysis of the consulting company Gartner, the
results of which are given in the previous part.
Scientists [30] reveal the phenomenon of business activation of AI technologies; meanwhile,
the authors emphasize the importance of an original combination of two established machine
learning algorithms (LDA and hierarchical clustering).
The right choice of Al technologies is crucial for every business structure. That is why the
matter of choosing a method that will ensure the selection of optimal hype AI technologies for a
specific business structure, taking into account its features, is becoming more relevant.
Analyzing the works of scientists in the field of problems, we see that the research singles out
the methods of classification [31, 32], data clustering [33], visualization [34], and other
approaches of AI, ML, and DP.
The conducted research made it possible to single out the methods that should be used in
practice to select promising Hype AI technologies for the organization. In particular:
- the analytic hierarchy process (AHP) − provides a systematic approach to decision-making,
which allows for structuring complex problems and taking into account many criteria when
evaluating alternatives;
- the Delphi method, which lies in experts giving their predictions about future technologies
or trends, and then summarizing the results. It is often used to predict technology trends;
- the method of scenarios − involves the development of various scenarios of technology
development, which allows the understanding of possible consequences of choosing a specific
direction;
- the SWOT analysis method, which makes it possible to assess the strengths and
weaknesses, opportunities, and threats associated with the implementation of specific
technologies.
- the technology forecasting method (Technology Forecasting) − uses the analysis of
historical data, development trends, and other methods to forecast the future development of
technologies.
- the method of startup scanning (Startup Scanning) − involves researching the market of
startups working in the field of artificial intelligence, which can help identify promising
technologies that are developing in this direction.
These methods can be used individually or combined to obtain a more objective and
complete map of promising Hype AI technologies for the organization. However, this research
will use the analytic hierarchy process (AHP), taking into account its characteristics.
Companies are developing strategies for implementing modern information and
communication technologies. The company’s innovative strategy in the IT field is formed
depending on the tasks to be solved taking into account market positioning, activity
specialization, expected competitive advantages, and the depth of crisis manifestations.
However, as research results [30] have shown, the use of artificial intelligence technology does
not have a significant impact on the effectiveness and efficiency of innovations. The results
suggest that in the era of artificial intelligence, the principle of “people-centeredness” in human
resource management and the acceptance of responsibilities by employees still play a very
important positive role [21]. In developing this issue, scientists [35] have pointed out that
artificial intelligence differs from other digital technologies, given its potential to become both a
general-purpose technology and a method of invention, and several companies are beginning to
integrate AI into their innovative processes.
To plan the company’s innovative strategy, first of all, it is necessary to establish the
most characteristic parameters, the value of which will determine the choice of a specific IT
project. The introduction of AI technologies is conditioned by the analysis of huge volumes of
data and the coordinated processing of multidirectional information flows at both technological
and content levels. Currently, the introduction of generative AI technologies, which lead to the
emergence of qualitatively new market opportunities due to the creation of a unique offer, has
intensified. However, it is extremely difficult to assess the real prospects of the emergence of
hype AI technologies. Solving this task requires a systematic approach, the necessity of which is
determined by the following factors:
• complexity of implementing hype AI technologies:
• multifaceted interpretation of such an innovation (infrastructure, process, and result);
• considering hype AI technologies as a whole “man-infrastructure-innovation”;
• the presence of a set of alternative innovations;
• the presence of a set of criteria for assessing the readiness of the environment through
the analysis of Hype Cycles for AI;
• the presence of a set of criteria for assessing the socio-economic effect of the
introduction of hype AI technologies;
• a lack of standards for the use of new hype AI technologies;
• subjective dependence of a set of innovation assessment criteria on the composition
and competencies of the expert group;
• the presence of unmanaged or poorly managed technological, socio-economic, legal,
and cultural factors in the system “man-infrastructure-innovation”.
The evaluation of hype AI technologies can be carried out by determining the socio-
economic effect [36], quality, or commercial potential. The above methods are reduced to an
expert assessment carried out by a survey or group examination through the following
approaches:
• determination of the socio-economic effect is carried out to evaluate an innovative idea or a
corresponding project;
• quality assessment comes down to measuring the consumer value of an idea;
• measurement of the commercial potential is carried out to assess the possibility of
innovation commercialization.
Such approaches are used to compare hype AI technologies to choose the best alternative for
the company. The formalization of the system includes formalization, structuring, its elements,
and interactions. Thus, the set determines the formal description of the company as a
system, where S = {s1, s2, ., sn} is a set of constituent elements of the system, C ={с1, с2, ., сm} is
a set of system connections and relationships between them. The task of implementing hype AI
technologies is defined as the process of transforming the company into the desired state. Thus,
the set determines control in the system, where O is a finite set of possible operators
that transfer the system from one state to another, and R is a desired state of the system.
Let the authors consider the possibility of applying the analytic hierarchy process for expert
assessment of the efficiency of new hype AI technologies.
To evaluate the implementation of hype AI technologies, it is advisable to use the analytic
hierarchy process, maximin convolution based on the theory of fuzzy large quantities, and the
method of analytical networks. All three methods are based on the formation of matrices of
pairwise comparisons of alternatives and criteria using the scale developed by T. Saati [37].
The research of the problem [38-41] made it possible to form the stages of applying the
analytic hierarchy process to ensure the selection of promising Hype AI technologies for the
organization
In the first stage, the authors will present a hierarchical decision-making structure regarding
the choice of hype AI technologies for the company (Fig. 2).
Socio-economic effect of the introduction of new AI technologies in the
Goal
organization
Criteria of the Interaction
Organization Informational Cost of
socio-economic effect with internal and
of personnel security ownership
of external
work
I level subscribers
Criteria of AI Preparedness Development
Coherence of Possibility of
readiness for the of the working of the network
AI technology customization
implementation of environment infrastructure
II level
Project 1 Project 2 Project 3
Alternatives ( AI product, (AI product, (AI product,
implementation implementation implementation
time) time) time)
Figure 2: Hierarchy of the influence on the socio-economic effect of the introduction of hype AI
technologies in the company
Source: developed by the authors
Further, based on the data of the company, the authors build matrices of pairwise
comparisons taking into account the criterion of socio-economic effect. When building
comparison matrices for the criteria of AI readiness, the authors take into account the results of
the Hype Cycles for AI analysis in Tables 3. At the next stages of the analytic hierarchy process,
the authors perform calculations for each variant of the coordination of elements of two levels
of criteria. At the same time, it should be noted that in the given approach some matrices will be
inconsistent. In this case, at the final stage of the hierarchy synthesis task, the homogeneity of
the entire hierarchy is evaluated by summing the homogeneity indicators of all levels, brought
by weighting to the first hierarchical level.
The application of this method allows for choosing the most effective project for the
introduction.
4. Conclusions
A strategy for analyzing projects of the implementation of hype AI technologies and the
most efficient means to ensure the effective development of domestic companies is elaborated.
In general, the following provisions are disclosed in the work:
I. The average annual growth of world investments in the AI field is more than 25%;
II. AI technologies are implemented in an increasingly wide range of business processes,
resulting in the formation of hype technologies;
III. The progress in the AI industry is researched by many international companies, in
particular, Gartner;
IV. The consulting company Gartner from 2019 to 2023 prepares annual reports in the
form of Hype Cycles for Artificial Intelligence, which show the high dynamics of changes;
V. The innovative strategy of the development of domestic companies, taking into
account progress in the AI industry is developed in conditions of high uncertainty;
VI. The task of the optimal development of domestic companies can be reduced to the
problem of fuzzy logic, for the solution of which a modified method of the analytic hierarchy
process can be used.
This method of determining the feasibility of projects of implementing hype AI
technologies can be perceived as a primary assessment. It allows for assessing the rationality of
the company’s measures as a whole, finding out according to which parameters it is necessary
to carry out active innovations. Next, more precise economic, technological, and marketing
justifications for specific innovative measures are needed, as well as the calculation of their cost,
yield, profitability, and resource availability.
5. Acknowledgements
The authors are thankful to all of the experts who have deeply participated into the study,
and express gratitude to reviewers whose insightful comments and suggestions have
significantly helped to improve the quality of the paper.
References
[1] Artificial Intelligence Market Worth 190.61 Billion USD by 2025 (prnewswire.com) URL:
https://www.prnewswire.com/news-releases/artificial-intelligence-market-worth-19061-
billion-usd-by-2025-674053943.html?ref=hackernoon.com
[2] Artificial Intelligence Market by Offering (Hardware, Software, Services), Technology
(Machine Learning, Natural Language Processing, Context-Aware Computing, Computer
Vision), End-User Industry, and Geography - Global Forecast to 2030. URL:
https://www.marketsandmarkets.com/Market-Reports/artificial-intelligence-market-
74851580.html
[3] M. Gonzalez, IMBA Forecasts a 15-fold revenue increase for the AI industry, with revenues
expected to rise from $18 billion in 2022 to $420 billion in 2027. URL:
https://www.linkedin.com/pulse/forecasts-15-fold-revenue-increase-ai-industry-rise-
gonzalez-imba-hdsme
[4] GenAI Solutions Help European AI Market Thrive in an Uncertain Economic Environment,
Says IDC. URL: https://www.idc.com/getdoc.jsp?containerId=prEUR251222623
[5] B. Gates: AI could kill Google and Amazon. URL: https://www.ghacks.net/2023/05/24/bill-
gates-ai-could-kill-google-and-amazon/
[6] How Businesses Are Using Artificial Intelligence In 2024 – Forbes Advisor. URL:
https://www.forbes.com/advisor/business/software/ai-in-business/#embed
[7] N. Shpak, A. Karpyak, O. Rybytska, M. Gvozd, W. Sroka, Assessing the business models of
Ukrainian IT companies. Forum Scientiae Oeconomia. – 2023. – Vol. 11, No. 1. – P. 13–48.
DOI: https://doi.org/10.23762/FSO_VOL11_NO1_2
[8] N. Shpak, L. Maznyk , Z. Dvulit , K. Doroshkevych , N. Horbal , S. Kis, Smart contract as a way
to exchange digital values in blockchain. Computer science and information technologies :
proceedings of IEEE 16th International conference CSIT 2021 (Lviv, Ukraine, 22–
25 September, 2021). – 2021. – C. 403–406.
[9] IDC FutureScape: Worldwide IT Industry 2024 Predictions (IDC #US50435423) URL:
https://www.idc.com/getdoc.jsp?containerId=US50435423
[10] M. Lee, H. Scheepers, A. Lui, E. Ngai, The implementation of artificial intelligence in
organizations: A systematic literature review. Information & Management. Volume 60,
Issue 5.2023.URL:
https://www.sciencedirect.com/science/article/abs/pii/S0378720623000642
[11] Y. Duan, Artificial intelligence for decision making in the era of Big Data–evolution,
challenges and research agenda. International Journal of Information Management. 2019.
URL: https://www.sciencedirect.com/ science/article/abs/pii/S0268401219300581
[12] Cardillo, How Many Companies Use AI? 2023. URL:
https://explodingtopics.com/blog/companies-using-AI
[13] L. Nguyen, Advantages of Artificial Intelligence (AI) in Business. 2023. URL:
https://revenuegrid.com/blog/advantages-of-artificial-intelligence/
[14] How companies use artificial intelligence. 2023. URL: https://www.eiu.com/n/how-
companies-use-artificial-intelligence/
[15] How is AI used in business in 2024? 2023. URL: https://chatfuel.com/blog/ai-in-business
[16] D. Spizheva, Using AI: Examples of How to Optimize Artificial Intelligence in Business. 2023
URL: https://turnkeystaffing.com/tech-trends/businesses-using-ai/
[17] Use of artificial intelligence in business. URL:
https://avadamedia.ua/ua/services/ispolzovaniye-iskusstvennogo-intellekta-vbiznese/
[18] Artificial intelligence for business: what tasks it can solve and in which industries it helps.
URL: https://mind.ua/publications/20254126- shtuchnij-intelekt-dlya-biznesu-yaki-
zavdannya-zdaten-virishuvatita-v-yakih-galuzyah-dopomagae
[19] P. Nitiéma, Artificial Intelligence in Medicine: Text Mining of Health Care Workers’
Opinions. J Med Internet Res 2023; URL: https://www.jmir.org/2023/1/e41138
[20] N. Chukhray, N.,Shakhovska, O. Mrykhina, L. Lisovska, I. Izonin. Stacking Machine Learning
Model for the Assessment of R&D Product’s Readiness and Method for Its Cost Estimation.
Mathematics., 2022, 10(9), 1466.
[21] Bi Siguo, Cong Wang, Bochun Wu, Shuyan Hu, Wutao Huang, Wei Ni, Yi Gong, Xin Wang, A
comprehensive survey on applications of AI technologies to failure analysis of industrial
systems. Engineering Failure Analysis, Volume 148, 2023, 107172,
https://doi.org/10.1016/j.engfailanal.2023.107172
[22] Hype Cycle for Artificial Intelligence, 2016. URL:
https://www.gartner.com/imagesrv/newsroom/images/emerging-tech-hc-2016.png
[23] Hype Cycle for Artificial Intelligence, 2019. URL:
https://emt.gartnerweb.com/ngw/globalassets/en/articles/images/top-trends-on-the-
gartner-hype-cycle-for-artificial-intelligence--2019-0.png
[24] Hype Cycle for Artificial Intelligence, 2020. URL:
https://emt.gartnerweb.com/ngw/globalassets/en/articles/images/2-trends-on-the-
gartner-hype-cycle-for-artificial-intelligence--2020-0.png
[25] Hype Cycle for Artificial Intelligence, 2021. URL:
https://emt.gartnerweb.com/ngw/globalassets/en/articles/images/the-4-trends-that-
prevail-on-the-gartner-hype-cycle-for-ai--2021-0.png
[26] Hype Cycle for Artificial Intelligence, 2022. URL:
https://emt.gartnerweb.com/ngw/globalassets/en/articles/images/hype-cycle-for-
artificial-intelligence-2022.png
[27] Hype Cycle for Artificial Intelligence, 2023. URL:
https://emt.gartnerweb.com/ngw/globalassets/en/articles/images/hype-cycle-for-
artificial-intelligence-2023.png
[28] Gartner Hype Cycle, 2023. URL:
https://www.gartner.com/en/research/methodologies/gartner-hype-cycle
[29] R. Mukhamediev, Y. Popova, Y. Kuchin, E. Zaitseva, A. Kalimoldayev, A. Symagulov, V.
Levashenko, F. Abdoldina, V. Gopejenko, K. Yakunin, et al. Review of Artificial Intelligence
and Machine Learning Technologies: Classification, Restrictions, Opportunities and
Challenges. Mathematics. 2022; 10(15):2552. https://doi.org/10.3390/math10152552
[30] A. Sestino & A. Mauro, Leveraging Artificial Intelligence in Business: Implications,
Applications and Methods, Technology Analysis & Strategic Management, 2022 34:1, 16-29,
DOI: 10.1080/09537325.2021.1883583
[31] Y. Kuchin, R. Mukhamediev, K. Yakunin, J. Grundspenkis, A. Symagulov, Assessing the
impact of expert labelling of training data on the quality of automatic classification of
lithological groups using artificial neural networks. Appl. Comput. Syst. 2020, 25, 145–152.
[32] N. Shpak, M. Odrekhivskyi, K. Doroshkevych, W. Sroka, Simulation of Innovative Systems
under Industry 4.0 Conditions. 2019. Soc. Sci. 8, 202. DOI: 10.3390/socsci8070202.
[33] T. Hastie, R. Tibshirani, J. Friedman, Unsupervised learning. In The Elements of Statistical
Learning; Springer: Berlin/Heidelberg, Germany, 2009; pp. 485–585.
[34] Van der Maaten, G. Hinton, Visualizing data using t-SNE. J. Mach. Learn. Res. 2008, 9, 2579–
2605.
[35] P. Hutchinson, Reinventing Innovation Management: The Impact of Self-Innovating
Artificial Intelligence, in IEEE Transactions on Engineering Management, vol. 68, no. 2, pp.
628-639, April 2021, doi: 10.1109/TEM.2020.2977222
[36] J. Wang, Z. Xing, & R. Zhang, AI technology application and employee responsibility.
Humanit Soc Sci Commun 10, 2023 p.356. URL: https://doi.org/10.1057/s41599-023-
01843-3
[37] T.L. Saaty, K. Kearns, Analytical Planning; The Organization of Systems, Oxford: Pergamon
Press 1985
[38] D. Margasov, E. Sakhno, I. Skeeter, Development of a model and modification of the method
of analysis of hierarchies for energy efficiency assessment. Eastern European journal of
advanced technologies. 2015. No. 5(2). - pp. 26-32.
[39] V. Vitlinsky Economic risk: game models. Kiyv, KNEU, 2002. 446 p.
[40] V. Dubrovin, V. Kruglikova, K. Fandeeva, The method of analyzing hierarchies in increasing
the effectiveness of an advertising campaign. State and regions - 2011. - No. 4. - P. 87-91.
[41] T. Romanchyk, Determination of the competitiveness of industrial enterprise projects
based on the analysis of hierarchies of their partial priorities. Communal management of
cities: science and technology. coll. 2014. No. 113. P. 95–101.