INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE
ISSN: 2692-5206, Impact Factor: 12,23
American Academic publishers, volume 05, issue 03,2025
Journal:
https://www.academicpublishers.org/journals/index.php/ijai
page 1687
ARTIFICIAL INTELLIGENCE AS A WAY TO CREATE NEW CONTENT
Suyunova Kamilla Baxromovna
Associate Professor of Marketing, PhD, Samarkand Institute of Economics and Service
Tulanbayev Akbarjon Akramjanovich
Third year student at Samarkand Institute of Economics and Service
Abstract
: Nowadays, artificial intelligence is becoming more widespread and significant in
various spheres of human activity. In this regard, many questions arise about the impact of
artificial intelligence on business processes, which areas of activity are most susceptible to
change. The article considers a number of areas in which artificial intelligence is actively used,
the tasks it can handle, as well as several neural networks that are widely used in 2025. The
work conducted a sociological study to identify the distinctive features of work with content
performed by neural networks and people.
Keywords:
artificial intelligence, neural network, automation, content generation, optimization
algorithms, automatic learning.
Introduction
The modern world is changing very quickly, and every year it becomes more complex. This
is facilitated by both the aggravation of military, racial and ethnic conflicts, the fight for gender
equality and universal tolerance, and technological phenomena: the introduction of high
technologies into almost all spheres of human life, rapidly spreading robotics, work with big data,
augmented and virtual reality and much more. Definitely, one of the main topics in the world of
technology in 2022 was artificial intelligence and its application in our everyday and professional
lives.
Analysis of literature on the topic
Artificial intelligence (AI) is revolutionizing content creation by leveraging neural networks,
machine learning, and automation. The Encyclopedia Britannica provides foundational
knowledge on AI technologies and their applications [1]. AI-driven solutions like ruDALL-E [4]
and Midjourney [7] generate visual content, while ChatGPT [6] assists in text generation. AI-
powered tools such as Yandex’s Balaboba [3] and Alice [8] enhance natural language processing
for interactive content. Moreover, AI is being integrated into branding and design, as seen in the
Logomachine project [5]. These technologies demonstrate AI’s role in streamlining creative
processes and making content generation more efficient and accessible.
Research methodology
This study employs a mixed-methods approach, combining qualitative and quantitative
analysis. A literature review was conducted using sources on AI-generated content, including
INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE
ISSN: 2692-5206, Impact Factor: 12,23
American Academic publishers, volume 05, issue 03,2025
Journal:
https://www.academicpublishers.org/journals/index.php/ijai
page 1688
text, images, and design. A comparative analysis was performed to evaluate AI tools like
ChatGPT, ruDALL-E, and Midjourney against human-created content. Additionally, a survey
was conducted with 125 participants to assess their ability to distinguish between AI-generated
and human-made works. Data analysis included statistical evaluation of responses and content
analysis of participant feedback. Ethical considerations were maintained by ensuring anonymity
and informed consent.
Analysis and results.
Artificial intelligence (hereinafter AI) is a set of technologies and methods that allow
computers to perform tasks that usually require human intellectual abilities, such as pattern
recognition, data classification, natural language processing, and many others [1]. One of the
main technologies used in AI is neural networks, which imitate the work of neurons in the brain
and allow computers to learn from experience.
The development of AI has a significant impact on various areas of life, including the labor
market. Every year, more and more companies are implementing AI to optimize business
processes, improve productivity, and reduce costs.
Improvements in technology and capabilities for collecting and processing data have allowed AI
to be used in many industries, from medicine and finance to marketing and design. Among all
the areas of activity that use AI, consider how it is used in the following areas:
Marketing. AI allows you to collect and analyze a large amount of data, which makes it
possible to optimize advertising campaigns and personalize communication with customers.
Machine learning algorithms are used to predict customer behavior and determine the most
effective channels for promoting goods and services. One example of such systems in Russia
is Yandex.Market, which uses neural networks for personalized product and service
recommendations. In 2022, at YaC 2022, a neural network named Ekaterina was announced,
named after one of the stylists working on the project, which was trained using data from
professional stylists. With its help, Market improved its search results for the clothing and
footwear category. The new technology allows customers to see fashionable items, regardless
of brand and price. Previously, due to the large number of products in the assortment, the
service did not have enough data on how exactly customers search for clothes and shoes. Now
Market will be able to improve its search results in other product categories as well.[2]
Working with text. AI can process and analyze texts in natural language, which allows
you to automatically generate product descriptions, create texts for websites and advertising
materials, and analyze the sentiment of reviews and comments. For example, the service
"Balaboba", created by Yandex, whichallows you to generate texts in Russian using AI. It was
launched in 2017 and has since become popular among users looking for a simple and
convenient way to generate texts. One of the main features of the service is its adaptability, it
takes into account the context in which the request was made and provides the most relevant
answer. [3]
Design. AI can automate the design process. AI can also be used to analyze designs and
identify the most effective elements that attract users' attention. In January 2021, OpenAI
introduced the DALL-E neural network, created to generate an image based on a text query in
English. In November of the same year, SberDevices and SberAI introduced the ruDALL-E
neural network, capable of working with a text query in Russian. At the moment, the neural
network has several generative models: ruDALL-E Kandinsky2.0 works in applications "Salut"
and Discord and creates images using voice commands; ruDALL-E Malevich is similar to the
INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE
ISSN: 2692-5206, Impact Factor: 12,23
American Academic publishers, volume 05, issue 03,2025
Journal:
https://www.academicpublishers.org/journals/index.php/ijai
page 1689
original DALL-E neural network, used by entering text queries in Russian; ruDALL-E Emojich
creates new emoji based on text descriptions, which can be used as stickers, clip art and
prototypes. [4]
Project Management: AI can help plan and manage projects using algorithms for
schedule optimization, risk forecasting, and budget management. AI can also help with
automated data collection and reporting, simplifying the project management process.
Neural networks have the ability to automatically learn and make decisions based on data,
which allows them to significantly facilitate operations and processes in the areas under
consideration, freeing people from certain tasks.
Since neural networks are actively used in the field of marketing. For creations individual
offers and consumer demand analysis, they can
replace human routine
tasks related to
providing recommendations, analysi sdata and forecasting the results of marketing campaigns
based on data from previous campaigns. Neural networks used to work with text, including
myself automatic content creation, language translation, sentiment analysisand keyword
identification can easily replace a person in performing tasks such as writing news or technical
texts, project descriptions, etc.
In design, neural networks can be used to create design layouts, generate ideas, and perform
automatic editing. Some routine design tasks, such as creating templates for websites or logos,
can be performed by neural networks with minimal human involvement. For example, Lebedev
Design Studio presented a new logo that combined the work of Logomashina studio designers
and neural networks. The idea is to create unique logos that can be generated automatically using
neural networks and then edited by designers to achieve the desired result. This approach can
significantly speed up the logo creation process and reduce the time and resources spent on
development. [5]
A number of project management tasks such as creatingWork schedules, budget planning or
risk assessment can also be performed without human intervention by neural networks based on
data analysis. In general, neural networks can replace human work in tasks that are structured
and repetitive in nature.
The history of machine learning development began in the middle of the last century, but one
of the main stages in the development of neural networks occurred in the mid-2000s, when
Geoffrey Hinton created new learning algorithms that allow efficient processing of large
amounts of data and training of deep neural networks with many layers. One of the main reasons
for the growing interest in neural networks was the improvement of computing power of
computers and the availability of large amounts of data. In addition, new tools for creating,
training and deploying neural networks have emerged, which have made this technology more
accessible to a wide range of people. Today, neural networks are widely used in various fields,
such as speech recognition, photo and video processing, and many others. Let's consider several
of the most popular neural networks in Russia in 2023.
ChatGPT is a powerful neural network model based on the GPT (Generative Pre-trained
Transformer) architecture that is trained on huge amounts of text data. It can generate texts in
more than 90 languages, answer questions, translate texts into other languages, create image
descriptions, and perform many other natural language tasks. ChatGPT works on the principle of
automatic learning, the model does not require pre-training for a specific task. Instead, it is
trained on a large amount of texts available in the public domain, after which it can be applied to
INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE
ISSN: 2692-5206, Impact Factor: 12,23
American Academic publishers, volume 05, issue 03,2025
Journal:
https://www.academicpublishers.org/journals/index.php/ijai
page 1690
solve various natural language problems.
One of the features of ChatGPT is that it capable process longtext sequences using technology
transformers, which allows the model to effectivelyanalyze and remember the relationships
between words in a sentence. Overall, ChatGPT is a natural language processing tool that can be
applied in many areas related to text processing.[6]
Midjourney is an AI-powered platform that was created to explore new means of thought and
expand the power of people's imagination. The platform's main feature is an interactive bot in the
Discord app that uses machine learning to generate images based on text entered by the user in
English.
This AI system uses deep neural networks to process text information and transform it into
visual images. A user can enter a text phrase into the bot that they want to illustrate, and
Midjourney will generate an image that displays the idea presented in the text. The Midjourney
neural network uses generative adversarial networks, which consist of two parts: a generator and
a discriminator. The generator creates an image, and the discriminator determines how realistic it
is. These two parts of the neural network are trained to interact with each other and improve their
results over time. One of the main advantages of Midjourney is that it can be used to create
unique, creative images that can help visualize ideas and concepts. This neural network can be
useful in various fields, including design, art, advertising, science, and technology [7].
Voice assistant Alice is a voice assistant developed by Yandex. Alice is powered by AI,
which allows her to recognize natural speech, answer questions, and perform various tasks using
voice commands. She is available on various devices, including smartphones, tablets, computers,
and cars.
The main functions of the voice assistant include: searching for information on the Internet;
reading weather forecasts and exchange rates; reading the main news; recognizing images in
photos; recognizing songs by short passages; ordering food; the ability to play games; ordering a
taxi; managing a smart home; plotting a route in the navigator; opening applications and much
more.
One of Alice's main advantages is that it is trained to work with natural speech, so users can
ask questions in natural language without using special commands or phrases. In addition, it has
the ability to understand context and take into account previous conversations with the user,
which makes it more intuitive and easy to use.
Overall, Alice is a convenient and innovative tool that helps users in their daily tasks and
provides ease of use in various areas of life.[8]
Many people believe that neural networks cannot yet replace humans in certain areas of
professional activity, especially when it comes to creative processes. Based on the material we
studied, we formulated a hypothesis: “Despite the development and adaptability of neural
networks, they can easily replace humans when performing structured work, but they cannot
replace the creative component of work. The use of such concepts as humor, complex metaphors
and allegories is not yet available to neural networks. A person can easily distinguish the work of
artificial intelligence from the results of human intellectual activity.”
In order to test this hypothesis, as well as to identify specific distinguishing features that a
person relies on when working with this or that type of content, we conducted an online survey
in which respondents had to choose from two texts and two images, which of them were
INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE
ISSN: 2692-5206, Impact Factor: 12,23
American Academic publishers, volume 05, issue 03,2025
Journal:
https://www.academicpublishers.org/journals/index.php/ijai
page 1691
generated by neural networks and which were made by a person. The survey involved 125
respondents aged 18 to 50.
The survey itself consisted of:
Text No.1 – generated by a neural network on request
"Write 5 sentences in a fictional style about how the feeling of love is born."
"In the depths of the soul, as in the most mysterious and dark corner, an extraordinary feeling
awakens, which smoothly flows into life energy, leading to a wonderful feeling - love. It begins
to grow and develop, like a tiny flower bud that penetrates into the very depths of the heart and
fills it with warmth and tenderness. Every day, every moment of life, this bud blossoms more and
more, filling the soul with light and spiritual well-being. And finally, when it reaches its peak,
the most beautiful flower appears in it - love, which gives happiness and joy to everyone
around."
Text No. 2 – an excerpt from the book “Genius and the Goddess” by O. Huxley.
"There is a saturated solution of feelings, and the reason for its crystallization can arise both
inside and outside. Words and events fall into this psychophysical chatter, and in it clots of
emotions and experiences are formed, calling for action. Then the glands develop, which leads to
the appearance of those very charming little animals of which the child is so proud and which
embarrass him so much. The solution of feelings is enriched with a new type of sensations, they
penetrate from the nipples, through the skin and nerve endings into the soul, into the
subconscious, into the superconscious, into the realm of the spirit. And these new centers of
mental tension of the personality seem to impart movement to the solution of feeling, make it
flow in a certain direction - to the absolutely unknown, full of mysteries sphere of love."
Image No. 1 – drawn by artist Maria Vyshinskaya. (Fig 1.)
INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE
ISSN: 2692-5206, Impact Factor: 12,23
American Academic publishers, volume 05, issue 03,2025
Journal:
https://www.academicpublishers.org/journals/index.php/ijai
page 1692
Fig. 1 "Drawing by Maria Vyshinskaya"
Image #2 – generated by a neural network based on the query “Night, cartoonish good
mythical creatures dancing in a circle around a fire.” (Fig. 2)
Fig. 2 “Image generated by a neural network”
25.6% of respondents completely correctly identified the work of neural networks and
humans. The main difficulty in the survey for respondents was identifying the texts, 21.6% noted
that both texts were written by a neural network, which in the first case was the correct answer
(Fig. 3), and in the second - not (Fig. 4).
Fig. 3 “Result of the survey on the selection of text written by a neural network”
Fig. 4 “Result of the survey on the selection of text written by a person”
At the same time, respondents who correctly marked both texts identified a number of
features that helped them in their choice:
INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE
ISSN: 2692-5206, Impact Factor: 12,23
American Academic publishers, volume 05, issue 03,2025
Journal:
https://www.academicpublishers.org/journals/index.php/ijai
page 1693
regarding the text written by the neural network, 8% noted that the narrative was too
simple, 6.4% noted that it lacked humor, 5.6% noted tautology in the text, 4.8% identified
speech, logical and punctuation errors, 1.6% did not see individuality in the text and noted
the feeling that the text was composed of works by other authors, and finally, 4% noted
that the text was “soulless”;
relative to human-written text 19.2
% highlighted the complexity of the narrative, 39.2% noted the use of literary devices, while
20.8% drew attention to the non-standard nature of the text, 6.4% mentioned the presence of
humor.
Based on the results of the survey on texts, it can be concluded that the non-standard nature
of the author's text, on the one hand, helped people distinguish it from the text written by the
neural network, but on the other hand, it greatly confused the respondents. At the same time,
regardless of the correctness of the answer, 13.6% noted the absence of humor in the text written
by the neural network, and 9.6% the presence of humor in the author's text.
64% of respondents successfully managed to select images. 92 respondents managed to
correctly identify the author's image, which amounted to 73.6% of the total number (Fig. 5),
while 44% of respondents singled out the author's style, and 32.8% noted that the human hand is
clearly visible.
Fig. 5 “Result of the survey on the choice of the author
images»
91 people coped with the image generated by the neural network – 72.8% (Fig. 6), 35.2% of
them noted that it was clearly noticeable that the image was generated by the neural network, 4%
paid attention to
inaccuracies in the image that allowed them to give the correct answer. Only 8.8% of
respondents decided that both images were generated by neural networks, and 9.6% thought that
both images were drawn by a person.
Fig. 6 “Result of the survey on the choice of an image generated by a neural network”
The results of the study show that modern neural networks are already developed enough to
create high-quality images and texts that are difficult to distinguish from human-created
drawings and texts. However, most people are able to understand that the generated images and
texts belong to neural networks due to the primitiveness of the created objects, errors and
inaccuracies. Writing text remains a difficult task for artificial intelligence, since human
language is a more complex area for modeling, since it has a higher level of abstraction and
contextuality than images.
Conclusions and suggestions.
INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE
ISSN: 2692-5206, Impact Factor: 12,23
American Academic publishers, volume 05, issue 03,2025
Journal:
https://www.academicpublishers.org/journals/index.php/ijai
page 1694
In conclusion, it can be added that the survey results highlight the importance of a
harmonious combination of human creativity and artificial intelligence in the modern world.
Despite the fact that neural networks can perform many tasks, in creative processes where
individuality, originality and creativity are important, humans are still irreplaceable.
The use of hybrid systems in which technical possibilities artificial-intelligence is used to
create an initial template, and then the person refines it and supplements it with their own ideas
and creativity, can be very effective, as it can help speed up and improve the creative process,
without replacing human creativity, but only complementing it and expanding its capabilities.
References:
1.
YaC, 2022. How to make the purchasing process predictable and teach algorithms to
understand
fashion Electronic resource]
(date accessed:
17.03.2023)
2.
Neural network "Balaboba". [Electronic resource]
https://yandex.ru/lab/yalm?style=0
(date
accessed: 17.03.2023)
3.
Neural network ruDALL-E. [Electronic resource]
(date accessed: 17.03.2023)
4.
Logo for Lebedev Studio: joint concept of Logomachine and neural network. [Electronic
resource]
(date accessed: 21.03.2023)
5.
Neural network ChatGPT. [Electronic resource]
Ihttps://openai.com/blog/chatgpt
(date
accessed: 21.03.2023)
6.
Midjorney neural network. [Electronic resource]
(date
accessed: 21.03.2023)
7.
Voice assistant "Alice" from Yandex. [Electronic resource]
(date
accessed: 21.03.2023)
