Specifics of Artificial Intelligence Application in Modern Media Space

Nerents Dar'ya Valer'evna PhD in Philology Associate professor, Department of Journalism, Russian State University for the Humanities 125993, Russia, Moskovskaya oblast', g. Moscow, Miusskaya ploshchad', 6, aud. 525 ya.newlevel@yandex.ru Other publications by this author

In 2023, the topic of the use of artificial intelligence and its consequences is increasingly relevant. Neural networks like ChatGPT-4 are becoming able to replace a person in the preparation of text materials, create photo illustrations on any topic, turn one person into another in videos using deepfake technology. All this requires deep reflection to understand the consequences of further development of AI and its impact on people's lives.

The most important data cluster that can have both a serious positive impact and become an extremely dangerous information threat is human data. Huge amounts of data are required for high accuracy of predictions and conclusions from pattern recognition algorithms. The more open a country is in terms of collecting personal data and the weaker the requirements for protecting the privacy of citizens, the more likely it is to create effective algorithms that can accurately guess the interests, preferences, habits, and desires of each user of digital technologies. Calls, online purchases, posts with photos or videos, geolocation marks on social networks, the use of smartwatches, mobile applications, etc. help to actively collect such data. All this is used by large IT giants, who, thanks to such an inexhaustible resource, are able to influence a person's worldview, life values, change a user's lifestyle and habits and contribute to the formation of public opinion.

Artificial intelligence is developing so dramatically that it is becoming less noticeable to humans. Personal data is being analyzed, clarified less and less obviously for the users themselves, and the human ability to understand how autonomous systems make decisions is decreasing. Thus, in 2023, it is safe to say that most of the audience is unable to understand how AI affects our lives, and this gap continues to grow.

The positive and negative sides of using AI have become the subject of lively discussions due to the realization that "machines" are no longer just assistants in solving the simplest tasks, but serious tools that are used in all spheres of life. Based on AI, they work:

- personal assistants (Siri from Apple, Google Assistant from Google, Alice from Yandex or Marusya from Mail.ru ), which are able to process voice commands, answer questions, control smart home devices, etc.;

- recommendation systems (YouTube or Ozon use AI algorithms to analyze customer or user preferences to make personal recommendations for movies, music, goods, services);

- autonomous cars (Tesla) with the help of AI are increasingly able to move in any road conditions in compliance with all traffic rules;

- medical diagnostics (AI is able to diagnose the disease using CT, MRI or X-ray images);

- machine translation systems (Google Translator or DeepL with the help of AI can translate texts with high accuracy into many languages);

- social networks (AI is able to personalize recommendations, filter content, both prevent and spread unwanted information at lightning speed);

- speech recognition systems (text transcription of audio and video files);

- face recognition systems (AI is able to compare images of faces with a huge database of photos, for example, to identify criminals).

The areas where AI occupies one of the important positions can be listed for a long time. These include agriculture (monitoring plant diseases, improving yields, monitoring equipment), meteorology (weather forecast), sports (analyzing game statistics, developing strategies, predicting the outcome of a match, etc.), finance (forecasting market trends, analyzing financial assets of companies, etc.), and law (systematization and analysis of certain court cases, maintaining standard reporting).

Despite the obvious advantages, it is important to understand that AI radically changes people's lives, making some actions easier and more accessible, while others are more difficult and incomprehensible. Such a leap in the development of neural networks does not allow humanity to realize and think about what is happening, forcing it to master and assimilate everything literally on the go (which is only worth the first book written by ChatGPT-4, called "Autobiography of a neural network", which was entirely written by AI using the data that was embedded in it). All this makes the modern user vulnerable from the point of view of media security, because human consciousness does not have time to isolate itself from all the innovations that manipulators actively use.

AI is able to create fakes of such high quality that they are indistinguishable from the original. In the future, such technologies, due to the creation of a false video or photo image, can not only ruin the life of one person, but lead to mass riots, rallies and even military clashes.

A sensational example of such possibilities was a fake photo of the Pope dressed in a fashionable down jacket and walking through the streets of New York, which was taken using the Midjorney neural network (Gosteva A. Viral photo of the pope in a stylish down jacket turned out to be fake // Lenta.ru . URL: https://lenta.ru/news/2023/03/27/thepope1 ). Or photos of the arrest of D. Trump, who is allegedly forcibly put into a police car (Makarychev M. Fake pictures of Trump's "forcible arrest" by police appeared on the Network // Rossiyskaya gazeta. URL: https://rg.ru/2023/03/21/v-seti-poiavilis-fejkovye-snimki-silovogo-aresta-trampa-policejskimi.html).

The so-called deepfakes have become especially dangerous in this context ^{[5, p. 74]}. Artificial intelligence combines a large number of photographic images and makes a video recording of them. The program can accurately determine how a person will react and behave in a certain situation. In other words, the essence of the technology lies in the fact that part of the algorithm studies the image of an object in detail and tries to recreate it, until the other part ceases to distinguish between the real image and the one created by the neural network. Deepfakes are almost impossible to recognize, since videos are usually characterized by a high degree of realism ^{[4, p. 35]}.

It is not uncommon for AI to be used to modulate voice and create a fake phone number, which allows you to deceive gullible citizens and extort large amounts of money from them.

Such examples demonstrate serious threats to an unprepared audience from AI. Manipulators and scammers pursuing their goals can skillfully use these "weapons". This issue became especially acute with the creation of the ChatGPT-4 neural network, texts, photos, videos of which it is sometimes impossible to recognize and identify generation. In the Russian Federation, the development of AI is treated cautiously, not seeking to introduce it in all areas of production abruptly and promptly.

However, this practice was particularly developed during the Covid-19 pandemic. In December 2020, Russian President Vladimir Putin, during a speech at the international conference on the future of artificial intelligence, noted the need for digital transformation throughout Russia in the next decade. Moreover, he stressed that thanks to the use of algorithms in the field of public services, it will soon be possible to solve each problem of citizens individually (Important statements were made by Vladimir Putin at the international conference on artificial intelligence // The first channel. URL: https://1tv-ru.turbopages.org/1tv.ru/s/news/2020-12-04/397930-vazhnye_zayavleniya_sdelal_vladimir_putin_na_mezhdunarodnoy_konferentsii_po_iskusstvennomu_intellektu). Since 2019, the use of AI has been discussed in more detail, which was largely facilitated by the appearance of the Presidential Decree "On the development of artificial intelligence in the Russian Federation" (Decree of the President of the Russian Federation dated October 10, 2019 No. 490 "On the development of artificial intelligence in the Russian Federation" // <url> URL: https://www.garant.ru/products/ipo/prime/doc/72738946 ).

In the media, the use of artificial intelligence at the present stage causes mixed reactions among journalists and other industry workers. The opacity and unpredictability of many of these technologies have caused widespread concern about their use in a wide variety of social spheres – from the justice system to media communications ^{[13, p. 14-15]}. In response, IT specialists, ethicists, sociologists and politicians are looking for ways to make algorithmic systems accessible, i.e. capable of meaningful understanding, including to ensure transparency and objectivity. These are important values for journalism, which strives for accuracy, impartiality and objectivity in order to gain the trust of the audience and maintain legitimacy in society. However, the question of what it means for AI and algorithms to be understandable in a journalistic context has not yet been fully resolved. A significant body of research indicates that these technologies are changing media practices and processes and reshaping the idea of what it means to create and consume news ^{[3; 7; 9; 10; 12; 14]}.

A number of Russian (V.N. Bogatyreva ^[1], A.O. Tretyakov, O.G. Filatova, D.V. Zhuk ^[8], A.D. Ivanov ^[4], D.V. Nerents ^[6]) and foreign researchers (R. Zamit ^[16], S. Lewis and M. Logan ^[11], J. Strey ^[15]), who note the instant search for requested information throughout the Internet space, scrupulous and accurate verification of factual information, creation of "template" news texts, monitoring of posts on social networks, analysis of data of any volume, identification of photo and video materials. So it is possible to note such positive aspects as saving time, facilitating the process of working with information retrieval and fact-checking, reducing financial costs (in the future).

The research conducted by the staff of the Faculty of Journalism of the Russian State University within the framework of the scientific project "Media Security in the era of digital transformations" is based on a continuous sample of materials from the Russian media (both federal and regional levels) in the period 2021-2023, designated by the authors themselves as texts or videos created using AI or machine learning. A total of 53 publications were analyzed. In addition, 10 in-depth interviews were conducted with representatives of editorial offices using AI in their work (RBC, Interfax, Rifei TV, 360) in order to identify the positive and negative sides of this activity.

The results suggest that as positive nuances, journalists note the opportunity to get away from routine tasks such as searching and processing information, selecting images, compiling background information, etc. As negative consequences, there are concerns about the reduction of reporters and news journalists, a large amount of funding that may not pay off, the constant availability of specialists whose work also needs to be paid, as well as tuition for the employees of the media enterprise themselves. In addition, there is no regulation of AI activities in the media either from the legislative or ethical side, which also makes such activities risky.

In 2022-2023, several regional TV channels launched presenters created by neural networks. On March 23, 2023, the news appeared that the 360 TV channel launched a weather forecast from artificial intelligence. GPT-4 technology generates a forecast for the day, and the Maxim bot reproduces the resulting data aloud (Will Artificial intelligence displace meteorologists? // Rossiyskaya Gazeta. URL: https://rg.ru/2023/03/22/iskusstvennyj-intellekt-vytesnit-meteorologov.html).

Thus, meteorologists first announce their forecast, and then broadcast the version from the neural network. And on the Stavropol TV channel Svoe TV, the virtual presenter has an already established image. Her name is Snezhana Tumanova, and she talks daily about the weather forecast on the TV channel. Several neural networks are working on her appearance at once: one is responsible for appearance, the second for facial expressions, the third generates text ("I'm not sick, I don't go on vacation": the girl created by neural networks became a presenter on Russian TV. What does she look like? // NGS24.ru. URL: https://ngs24.ru/text/entertainment/2023/03/23/72158123). The same practice was started on DON 24, launching a virtual presenter Aksinya to voice the weather forecast on the TV channel.

Literally the next day, the Permian TV channel Rifei-TV used a virtual image as the host of the news program Evening on Rifei. He voiced one of the news (The neural network became the news anchor on the Permian TV channel Rifey-TV // Rifey. URL: https://rifey.ru/news/list/id_122380). A few days later, the news appeared that a virtual presenter named Daria appeared on the Prima TV channel in Krasnoyarsk, who shared the news on the air and also asked several questions to the audience. The text for her was also generated by a neural network (The girl created by neural networks became a presenter on Krasnoyarsk TV // NGS24.ru . URL: https://ngs24.ru/text/job/2023/03/25/72164072 ).

At the same time, in addition to conducting TV projects, AI is used on Russian television to create screensavers (Khabarovsk), design posters based on project synopses (Yu), and edit news digests (Yamal-Media).

However, AI is used not only on television, but also in online publications, as well as in the press. An example of RBC is noteworthy, which published the issue of the newspaper (No. 61 dated 04/28/2023) together with GigaChat, a neural network created by Sberbank (RBC Gazeta. Issue No. 61 (3731) dated 04/28/2023. URL: https://www.rbc.ru/newspaper/2023/04/28 ). All materials relate in one way or another to the application of new technologies in various fields and fields of activity, in addition to the "traditional" text written by a journalist, materials from GigaChat are published in a separate insert on a light green background. These are mostly sammari (summaries) of written articles or background information, as well as generated illustrations.

And the Russian Media Group launched a radio station called Neuro Flow, created by the Mubert neural network. The neural network has already written about 200 tracks in the dance house format (Ignatiev D. Russian Media Group launched an AI-created radio station on the Internet // Vedomosti. URL: https://www.vedomosti.ru/media/articles/2023/03/28/968350-russkaya-mediagruppa-zapustila-v-internete-sozdannuyu-ii-radiostantsiyu).

However, experiments with the introduction of AI into the media environment began several years ago. So, back in 2017, the media holding Hearst Shkulev Media began using Jenny's algorithm to create journalistic texts (Kholina A. Sex machine: who is better in bed with - a robot or a human? // Elle. URL: https://www.elle.ru/otnosheniya/lubov-i-seks/seks-mashina-s-kem-luchshe-v-posteli-s-robotom-ili-chelovekom). Despite the fact that the first experiments were not very successful, about a year later, Elle magazine (ceased to be published in the territory of the Russian Federation) confidently began to publish materials written by a "robot". And in 2018, the domestic TV channel TNT4, focused on entertainment content, began using a neural network in order to study viewer interests and preferences within the framework of TV viewing.

Moreover, today AI has become a really important tool for data journalists working with Big Data. Processing large data sets, systematization, aggregation, generalization, and tabulation of data according to specified parameters are just a few tasks that computer algorithms are able to perform.

The analysis of media texts created with the help of AI demonstrates that the modern stage is characterized by a number of possibilities for its application:

- creation of automated (of the same type) news and advertising texts and reference materials without errors and typos;

- News monitoring (or tracking data on a specific topic);

- data collection and aggregation (including Big data);

- analysis of posts or photo images (videos), tweets, comments on social networks;

- communication with the audience via chatbots;

- creation of algorithms and special codes within the framework of information analysis (especially of a financial nature);

- fact checking (checking statistical indicators, dates, names, etc.).

The above advantages are direct evidence that the competent use of IT technologies can not only facilitate, but also increase the efficiency of the workflow in any media sphere.

The interviews resulted in conclusions regarding the nuances of using AI in editorial offices. Currently, there are many possibilities for using artificial intelligence in the media sector. Some AI systems are huge and capable of performing a large number of calculations, while others are narrow and designed to solve only one task.

8 out of 10 interviewed journalists indicated that their editorial offices use small neural networks capable of performing a number of routine operations that do not require significant energy consumption and data.

At the same time, the creation of fake content, intentionally or inadvertently, is considered a serious risk for the audience from the position of using AI in the media. Social networks and blogging content are a source of information not only for the youth audience, but also for many journalists who, in pursuit of traffic and popularity, seek not so much to check the news as to become the first and publish an "exclusive" (9 out of 10 interviewees called this one of the main problems). Thanks to AI, fakes (and even more often, deepfakes) become indistinguishable from real content, which is what detractors use, deceiving even experienced reporters.

Another option is when the neural network itself makes mistakes by incorrectly reading the designations (for example, instead of 1-7%, indicating 17% or instead of 1925, talking about 2025). Such factual errors in the news about stock quotes or financial transactions can lead to global consequences. The texts created by the neural network may lack the context of what is happening, which will also lead the reader to the wrong conclusion. This was stated by 7 out of 10 interviewed journalists.

Another important aspect from an ethical point of view is the lack of indication when the material was published by AI and when by a real journalist. For example, the news feed in many online publications is based on the principle of lack of authorship, which does not allow the reader to see who wrote this text and in what way. For example, a sports publication Sports.ru It uses AI when maintaining a sports chronicle and generating various headlines, which sometimes have errors and inaccuracies. And then the next question arises: who is responsible for inaccuracies, factual errors, misinterpretation of data or comments? Does the audience agree to read texts generated by neural networks or watch the news with virtual presenters?

The audience's distrust of neural networks and computer algorithms can become a significant obstacle to the development of AI in the media sphere as a whole. According to a VTsIOM study, in 2022, more than a third of Russians surveyed (32%) do not trust AI technologies. The strongest fears are the leakage of the data they collect, the use for selfish purposes and the risk of making decisions for which no one is responsible (The Russians named their main fears of artificial intelligence // RBC. URL: https://www.rbc.ru/society/28/12/2022/63ab45de9a7947664c3ef893). In addition to the above, respondents believe that AI will lead to the degradation of the population (which, in general, may be close to the truth, given recent trends in education, when students consider it acceptable to submit a text generated by neural networks as an abstract, course project, and even a WRC), as well as the risks of systematic failures and errors in work.

The lack of qualified personnel to set up and manage such software is also important (6 out of 10 interviewees named this problem). This does not allow the editorial board to go deeper and explore all the possibilities of the neural network, but, as they say, go "to the top" using only accessible and understandable functions. This approach leads to a low quality of the created content and a negative attitude of the audience towards such experiences. As a result, the editorial board loses a lot of money and cannot continue its experiments.

A significant difficulty is the lack of a complete picture of the world (algorithms create an "information bubble"), when it is the AI that decides for the user what is important and what is secondary and can be ignored, thereby preventing a person from seeing everything that is happening around, but focusing only on a small part. As a result, this leads to limited perception, and journalistic texts with this approach cannot claim to be a high-quality, full-fledged media product.

It is still impossible to say for sure whether the pros of using AI outweigh the cons. But in contrast to the negative consequences, positive measures can also be noted. For example, Roskomnadzor is discussing the prospect of using artificial intelligence to combat disinformation on the Internet. AI will be able to identify deepfakes, determine the context of what is happening on the video and verify the facts. Its implementation can take place in a period of 3 to 5 years (Gureeva Y. Roskomnadzor uses artificial intelligence in the fight against Internet fakes // Rossiyskaya gazeta. URL: https://rg.ru/2023/04/24/roskomnadzor-ispolzuet-iskusstvennyj-intellekt-v-borbe-s-internet-fejkami.html).

It can be noted that the conducted research suggests significant time and financial savings when using artificial intelligence technologies in media. This factor was mentioned by all 10 interviewed employees of the media sphere. In addition, everyone also noted the simplification of the work of journalists by performing routine tasks by the "machine", which in turn increases productivity.

AI is able to recognize patterns and make quick decisions, find patterns in large datasets and make accurate predictions. Artificial intelligence has evolved from a simple program to help you find information to a complex system capable of creating analytical texts followed by minimal editing by humans.

The negative consequences are primarily that AI can present inaccurate or incomplete information (for complete data, it is necessary to constantly replenish the database), misinterpret the context, and reinforce existing stereotypes and prejudices. In addition, short-term memory limits neural networks in their ability to process large texts, and also does not allow them to understand emotions and feelings in order to express them in the material in the future.

The corpus of analyzed publications demonstrates a clear difference between virtual and real authors, which consists in the richness of expressions used, the emotional coloring of the data presented, and the correct interpretation of events. The AI is able to present factual information, brief information, voice the prepared text, nothing more. At the same time, the continuous improvement of neural networks suggests in the near future the possibility of creating analytical materials created with the help of AI, which will be able to identify cause-and-effect relationships, draw conclusions, create accurate forecasts based on the data system loaded into it (7 out of 10 media representatives surveyed see such a future).

At the same time, there are concerns that the audience is not yet ready to see virtual presenters and robots instead of real journalists. Sociological research over the past two years shows that Russians rather negatively perceive AI as the creator of media content, being skeptical and extremely wary of machine texts. Questions of legal liability for inaccurate data and ethical regulation of the work of neural networks as authors or TV presenters remain open. All these important factors indicate the scientific potential of this topic and the prospects for further discussion.