Threads and upgrades of AI technology in the new media era

Traditionally, media organizations have had an inner hierarchy, and on each level of it there is a specific job representative. However, with the rapid development of technology, including introducing artificial intelligence into all aspects of modern society, the industries we are familiar with are becoming more and more mechanized. In recent years, AI algorithms have developed its collect and process large amounts of data ability, now AI technology are able to create content on its own and personalize it based on the user's data. Some leading media outlets are already used to using AI in their workflows. Nevertheless, there are many experts who remain sceptical about this trend of great acceptance of AI technology. So does AI serve as a progressive tool for a new era of media or is it just a potential threat? This is the question that this thesis attempts to answer. The thesis contains an exploration of the definition of AI, an analysis of the impact AI technology has on media organization and their employees, as well as an indication of the potential threats of AI.

Key words: artificial intelligence, automation, mechanization, new media, big data. artificial intelligence automation mechanisation

I. Definition of Artificial Intelligence and its Operational Features

Artificial intelligence is a phenomenon that emerged in the 1940s. There are no resources to say where the development of this field of science began, because computer machines have existed for a long time, but before that these machines did not have the characteristics of artificial intelligence. The term "Artificial Intelligence" was first used by the American scientist John McCarthy. At that time, the computer practitioner described "artificial intelligence technology" as "a technology that can simulate every aspect of learning or any feature of intelligence"⁰⁵].

In recent years, artificial intelligence has become a hotly debated phenomenon reaffirming that the goal humans have is to create a machine that can think like a human and/or surpass them. From the very beginning to now computer scientists hope that new AI technologies will be able to store larger amounts of data, make logical connections faster, and make long-term predictions based on available data.

The capabilities of artificial intelligence and the prospects for its development will most notably require continuous refinement of its capabilities to:

1) Logical thinking (enabling machines to perform complex tasks that people can perform)

2) Knowledge sharing (allowing machines to analyze and describe objects, people, and language)

3) Planning and navigation (enabling machines to get from point A to point B, as in the existing example - self-driving cars, which can independently establish and follow a route)

4) Natural language processing (enabling computers to understand and analyze speech, a function that is already actively used in speech recognition technology)

5) Perception (adapts the computer to interact with the world through sight, hearing, touch and smell)

6) Empirical intelligence (giving machines the intelligence to be able to simulate emotional intelligence, moral judgments, and other attributes inherent to humans) [15]

An artificial neural network is a system of simple processors that connect and interact. In fact, neural networks mimic the process of human neural activity and thinking, but the difference is that AI thinking is replaced by computation. The purpose of computation is to simulate human recognition processes and logical conclusions.

Artificial neural networks are used to solve complex problems, such as:

a) pattern recognition (text characters, images, sound samples, etc., which is currently the most widespread application area of AI techniques) ;

b) data distribution by parameters;

c) decision making and control;

d) clustering (dividing the input information into different categories and later learning to determine in advance to which category the input information belongs);

e) prediction;

f) data compression and associative memory^[15].

However, after this technique received some degree of acceptance, many different scholars are doubting its ability to fully simulate human cognitive abilities, and the main reason is that there is a fundamental difference between the brain architecture and the computer architecture. A traditional computer processes data very fast, but it mostly processes data sequentially according to well-defined algorithms. The animal or human brain works much slower than a computer, but can process large amounts of disparate data in parallel.

II. AI technology as a new media perfection tool

1. Optimizing the speed of collecting, analyzing and generating information

The Associated Press (AP) was one of the first organizations to use AI in the process of collecting, analyzing, and generating information. In 2015, Associated Press also announced that it would be applying AI to its quarterly earnings reports. With technology from Automated Insights Technologies, the company has been able to automate its quarterly earnings reports, increasing its productivity tenfold and generating more than 3,000 reports per quarter, compared to 300 previously^[6]. In addition, the reports generated by AI all contain a lot of information, such as company descriptions, expert opinions, all the necessary facts and figures, and most importantly the number of errors in the reports has been significantly reduced.

The Korean news agency Yonhap has created an AI system that can write out sports notes in a very short time. Soccerbot works on the basis of algorithms that mimic the work of field reporters^[4]. The process of preparing material by this intelligent system can be divided into three stages: collecting data, writing content, and checking grammar and spelling.

Now AI is able to perform some multi-class tasks in less time than humans use. For example, the Election DataBot from ProPublica collects a lot of information about the U.S. presidential election, makes real-time reports, describes polls, Google search trends, etc^[22]. The famous newspaper The Washington Post also uses AI technology for its daily work of gathering facts. Fox Broadcasting's AI automatically generates sports scripts that appear on the Big Ten Network website, while Yahoo uses similar technology to write sports reports for different types of fantasy sports, and is able to provide this information individually for each user^[25]. It is known that algorithms do not get tired or distracted, so the potential of AI to improve the speed, workload, and accuracy of work in the media industry is enormous.

2. The information filters

The fact checkers of many media platforms are namely AI machines. Due to their advanced search and recognition capabilities, AI technologies are often able to completely replace personnel credits in their tasks. In 2017, the Associated Press announced a project to use AI machine which performs "automatic image processing" and is capable of processing 3,000 to 4,000 images per day^[13]. This technology is able to recognize faces, facial emotions, ongoing actions, and colors. The purpose of this artificial technology is to improve the accuracy of the search and simplify the function of automatic search, in addition, it is able to automatically identify images with sensitive content, such as violence, nudity, cruelty, etc. The aforementioned Korean Soccerbot also collects data from five different sources during its work in order to identify inaccurate information^].

In 2015, Facebook started testing a new technology called DeepFace^[24]. The goal of this AI technology is to search users by photos. The technology works according to a complex algorithm: first, it distinguishes a person's face, determines about a hundred points with the help of an intelligent system, and creates a 3D mask that can be used to find a similar user's face based on this data. This technology is able to recognize people with 97.5% accuracy, and DeepFace works even if the user is sideways to the camera or has a different facial expression in another photo.

3. Improving interaction between platforms and users

New media influenced by AI technology can make new media platforms more efficient and communicative by placing a piece of content that needs to be highly promoted at the top of users' minds. Artificial intelligence systems are able to provide users with the most relevant material based on their data (cookies, cache, history, etc.). As a result, users will receive material and news that is as personalized as possible.

Jeremy Gilbert, director of strategic initiatives at the Washington Post, stated that the publication already uses AI for projects promoted in certain regions (promoting content by geography)^[5]. The New York Times announced that they have not published reports generated by AI, but have used AI systems for personalized content recommendations, moderating comments, and identifying images^[18].

III. The threat that AI technology brings to the new media world

Nowadays, AI systems are not only used to optimize a work process, but also as fraud tools. In the era of information technology, the means of deception machines have become more adapted to modern reality. The artificial technologies used, which can replace a face in a photo, replace a person in a video or change the sound in an audio file, have a direct impact on the work of media platforms, since a photo can be faked, a video can be edited and a sound can be changed causing the reliability of the information received to become more difficult to analyze.

The first software to fake photos on the basis of artificial intelligence algorithms was Reflect Faceswap, invented by Ukrainian computer researchers. This machine can replace the face of one person with the face of any other person: the face of a friend, a celebrity or a character from a painting, a game, a TV show, or even a sculpture. The special thing about this technology is that it not only replaces another face with another face, but is able to preserve the emotions and facial expressions of the person in the original photo. This is a revolutionary technology, as most graphic editors can only replace "masks", but cannot preserve facial expressions while giving the face new features. However, "Reflect Faceswap" imposes a mask change when it is able to be presented at a new picture with the exact same facial expression as in the original image.

First, this system determines the main points of the face in the first photo (eyebrows, lips, eyes, nose, width and height of the face, etc.) and then determines the same points in the image that will replace the original photo. Reflect Faceswap also uses the shadows on the face to analyze the characteristics of light and skin tones. Next a new photo is created by the image generator, which has new facial features but also the same facial expressions. This AI technique uses Generative Adversarial Network (GAN) algorithm in its work^[19].

Such an intelligent technology is a big step forward for many industries, but it can pose a huge threat to the media industry. The artificial intelligence systems learn and evolve extremely quickly, making it much harder to detect fake material. This means that the reliability of material received by media companies is increasingly questionable, and that every photo needs to be scrutinized, which will greatly complicate the business of media.

In addition, we have recently been able to see more and more so-called "digital creativity" created by AI machines, such as a Twitter account that parodies and posts Trump's message style. Sociological studies have shown that AI bots have the greatest impact on users when their messages are consistent with people's preexisted opinions. Trump supporters tend to respond to messages spread by proTrump's bots, which subtly model Trump's style and readily promote the messages they write^[21]. This means that modern AI can not only fake photos or videos, but also come out with false information material, model various sources, and falsify various data, all of which in turn complicate the work of the entire media industry.

Some governments are also questioning the consequences for the concept of democracy if the algorithms assume the oversight role of the media industry (most notably journalism) over the government^]. If AI technology will identify truth and falsehood, set public agendas, and control public opinion, it is a big question whether society can trust the machine algorithms it uses.

IV. AI technology's assistance and challenges for new media workers

1. AI as a tool to simplify the work of media professionals

Many new media experts believe that their industry will not become one of the fully automated industries. Some studies suggest that the current level of AI technology can only fully automate 15% of journalists' work as well as 9% of editorial work^[14]. Compared to AI, people still have an advantage in the media industry in some key factors, such as live communication, expert thinking, adaptability, and creativity.

Glenn Kessler, editor of the Washington Post fact-checking blog, argues this way: "The media industry still requires people to perform more complex tasks, such as detecting bias, but computers can greatly help monitor and identify information worthy of personnel attention. It is a personal assistant that points us to information that we may have missed"^[23]. The impact of using AI simplifies the work of media workers in two main ways.

1) reduces the amount of work involved in generating content. AI systems have now learned to write information genres, while people can spend their time on more complex and interesting tasks that will affect the quality of the entire institution.

2) allows people to avoid the work of reviewing large amounts of information data. Because intelligent systems have the ability to categorize text, images or other content by topic, media workers do not have to look at large amounts of such information. The machine is able to do it independently, while personnel only need to select the necessary material for their work.

Bloomberg, one of the leading providers of financial information for the financial markets, uses the artificial intelligence system Cyborg for content generation and management. This system helps media workers to provide thousands of reports on company earnings each quarter. The technology analyzes a particular company's financial report the moment it appears and writes a business information note containing the most important facts and figures^[18].

2. The potential of AI technology as a replacement for new media workers

The development of AI technologies will inevitably lead to the replacement of certain occupations in new media. Futurist Amy Webb predicts that within the next 10 years, occupations such as journalists, graphic designers, promoters, news feed editors, and fact checkers will disappear from newsrooms. Webb also suggests that the next step in the development of online media will be based on artificial intelligence technologies, that is, intelligent systems that are increasingly able to perform creative functions traditionally seen as the prerogative of humans^[9].

Nowadays news organizations are already using AI systems to collect and analyze data from multiple sources and combine these data into articles. For example, Forbes has developed an AI-based content management system called Bertie, an AI machine that identifies newsworthy information and suggests headlines for it on its own^[12]. The Washington Post invented Heliograf, which can use quantitative data to write entire articles. During the 2016 U.S. presidential election, the technology published more than 500 reports^[16]. For several years, the Olympic Games as well as the FIFA World Cup (including this year's) have been using AI technology to collect data with writing reports and even predicting the results, and publishing them on all official platforms^[10].

In addition, AI technologies are able to maintain 24-hour work in order to significantly improve the efficiency of media organizations. For example, Xinhua News Agency has a robot reporter named "І&Ж'МЛ" kuai bi xiao xin. It can automatically finish the editing and finishing work after the release of some news stories, and it can write and publish the articles within 24 hours. Or for instance, in 2018, Xinhua News Agency launched the world's first artificial intelligence news anchor, which replaces a lot of labor of real anchors and is able to broadcast continuously for 24 hours^[1]. At the same time, this technological breakthrough has given rise to many discussions, and the question is whether AI will disrupt the entire industry, leading to an underestimation of the demand for employees.

The trend mentioned in this thesis is the point that modern media workers are thinking about: what else can employees do when artificial intelligence collects and analyzes data by itself? American inventor Elon Musk has spoken out about the idea that AI machines will lead to mass unemployment. He called AI "the greatest threat to human civilization" at the National Governors Association Summit^[11]. Jeremy Gilbert, director of strategic initiatives at The Washington Post, supports another viewpoint. According to him, AI is involved not to replace personnel, but to help them focus on more complex tasks^[5]. A large number of scientists have stated that automated media is still imperfect. If we talk about journalism, then an article generated by AI technology can only answer the "what?" question and cannot yet explain the "why?"^[25].

Gu Zhouhao, editor of China's Zhejiang news client, also supports the view that AI cannot replace human thinking: "Some or many would say, in the era of big data, various platform portals have started to use algorithms to count readers' interests, a matter that can be left entirely to computers. But I think, a reader, who has read the car section today, does not sure that he still wants to continue to pay attention to the news about the car tomorrow. Perhaps the future development of technology will enable targeted personalized news to all readers through artificial intelligence, but it seems that for now, algorithms cannot replace judgment and AI cannot replace thinking"^[2]. He also mentions that if one does not have a deep understanding of a topic or phenomenon, one (AI) cannot perceive the meaning behind an event and decide its value.

Conclusion

Modern technologies, including artificial intelligence, have opened up many opportunities for human development, which can optimize various industries, and the media industry is one of them. Different experts have different views on the future of the industry. Some scientists believe that intelligent technology will become only an auxiliary tool for work, other scientists tend to believe that future artificial intelligence will gain a sense of cognition and competence and learn to complete tasks like humans do.

Mark Zuckerberg, founder of Facebook and chairman and CEO of Meta, and Elon Musk, founder of SpaceX, Tesla and PayPal, have also argued over this issue. Musk has called for greater U.S. regulation of artificial intelligence technology, warning that it poses a great threat to humanity. Zuckerberg, on the other hand, called these statements "irresponsible" and stated that he maintains an optimistic view of the future of AI technology^[11]. It is worth mentioning that Zuckerberg has already felt the negative features of AI himself. when Facebook launched chatbots, and was forced to shut it down the following year for the reason that these artificial bots made an attempt to communicate independently, when there was a chance to get out of the control of IT specialists^[3].

Artificial intelligence does have a negative impact on new media industry, due to the learning and rapidly improving capabilities of this computation, it is now increasingly difficult to identify fabricated information or the reliability of the source of information, leading to mislead of society and to users' misunderstanding.

However, we would describe the impact of AI on new media as progressive and innovative. Forbes, the Associated Press, The Washington Post, Xinhua and various other social media software use AI in order to optimize their operations and the work of their employees. While AI cannot yet fully perform the tasks of reporting, listening, and using creativity to create new material on its own, it can often simplify human work, thus taking the media industry to new level.

The new media workers will still be valuable in the analytical and argumentative fields of media, which have creative, emotional, and evaluative components. This means that media workers should understand fields such as sociology, political science, philosophy, and history in order to move from being simple statisticians to experts who can professionally analyze the social, national, and world events that occur. Today, these tasks are generally performed by professional philosophers, political scientists, economists, historians and sociologists, while new media workers are often unable to compete with them simply because of their lack of theoretical knowledge. In short, there is a tendency to mechanize all routine operations in the modern media industry, and this phenomenon can indeed threaten the disappearance of some media professions. At the same time, however, all tasks that require creativity and involve human values and emotions will still be the privilege of media workers.

References:

[1] Bradley, T. (2017). Facebook AI Creates Its Own Language In Creepy Preview Of

Our Potential Future. Forbes. Retrieved from

https://www.forbes.com/sites/tonybradley/2017/07/31/facebook-ai-creates-its- own-language-in-creepy-preview-of-our-potential-future/?sh=31d017292c05

[2] (LEAD) Yonhap, NCSOFT to jointly carry out AI research. Yonhap News Agency. Retrieved from https://en.yna.co.kr/view/AEN20180516003251320

[3] Chaudhuri, S. (2019). Robo-Reporting: How Artificial Intelligence Is Assisting Journalists in the Newsroom. The Mack Institute For Innovation Management.

Retrieved from https://mackinstitute.wharton.upenn.edu/2019/journalists- jeremy-gilbert-washington-post/

[4] Daly, C. (2017). Future ofJournalism Will Be Augmented Thanks to AI. AI Business.

Retrieved from https://aibusiness.com/verticals/future-of-journalism-will-be-

augmented-thanks-to-ai

[5] Deepfake video app Reface is just getting started on shapeshifting selfie culture. Lingomine. Retrieved from https://lingomine.io/library/31602

[6] Duberry, J. (2022). Artificial Intelligence and Democracy Risks and Promises of AI- Mediated Citizen-Government Relations. Edward Elgar Publishing Limited, 72-89.

[7] Edge A. (2016). Journalism jobs of the near future, according to Amy Webb. Journalism.co.uk. Retrieved from www.journalism.co.uk/news/journalismjobs- of-the-near-future-according-to-amy-webb-ona16/s2/a673836/

[8] Felton, J. (2022). AI Has Run 100,000 Simulations And Predicted The 2022 World Cup Winner. IFLScience. Retrieved from https://www.iflscience.com/ai-has-run- 100-000-simulations-and-predicted-the-2022-world-cup-winner-66314

[9] Hamblin, A. (2017). Why Elon Musk, Mark Zuckerberg artificial intelligence fight

is far from over. The San Diego Union-Tribune. Retrieved from

https://www.sandiegouniontribune.com/opinion/the-conversation/sd-elon- musk-mark-zuckerberg-feud-20170726-htmlstory.html

[10] Kalim, F. (2019) Forbes doubles monthly visitors with Bertie, an AI-driven CMS.

What's New in Publishing. Retrieved from

https://whatsnewinpublishing.com/forbes-doubles-monthly-visitors-with- bertie-an-ai-driven-cms/

[11] Leveraging AI to advance the power of facts. The Associated Press. Retrieved from https://www.ap.org/discover/artificial-intelligence.

[12] Manyika, J. , Chui, M. et al. (2017). Harnessing automation for a future that works. McKinsey & Company. Retrieved fromhttps://www.mckinsey.com/featured- insights/digital-disruption/harnessing-automation-for-a-future-that-works

[13] Moore, J. C. (2018) Artificial Intelligence: What Everyone Needs to Know Today About Our Future. CreateSpace Independent Publishing Platform.

[14] Moses, L. (2017). The Washington Post's robot reporter has published 850

articles in the past year. DIGIDAY. Retrieved from

https://digiday.com/media/washington-posts-robot-reporter-published-500- articles-last-year/

[15] Peiser, J. (2019). The Rise of the Robot Reporter. The New York Times. Retrieved

from https://www.nytimes.com/2019/02/05/business/media/artificial-

intelligence-journalism-robots.html

[16] Reflect Face Swap. BetaPage. Retrieved from

https://betapage.co/startup/reflect-face-swap

[17] Searle, John (1990, January). Is the Brain's Mind a Computer Program?. Scientific American, 262 (1), 26-31.

[18] SMIALEK, J. (2018). Twitter Bots May Have Boosted Donald Trump's Votes by

3.23%, Researchers Say. Time. Retrieved

from https://time.com/5286013/twitter-bots-donald-trump-votes/

[19] Willis D., LaMarr G.,LeMee (2018). The Election DataBot: Now Even Easier. ProPublica. Retrieved from https://www.propublica.org/nerds/the-election- databot-now-even-easier

[20] WNIP (2018). AI: An invaluable sixth sense for journalists?. What's New in Publishing. Retrieved from https://whatsnewinpublishing.com/ai-for-journalism- data-driven-sixth-sen

[21] WOOLLASTON, V. (2015) Facebook can tag you in photos AUTOMATICALLY:

Social network starts rolling out DeepFace recognition feature. MAILONLINE. Retrieved from https://www.dailymail.co.uk/sciencetech/article-

2946186/Facebook-soon-tag-photos- AUTOMATICALLY-Social-network-starts- rolling-DeepFace-feature.html

[22] Сухачева, А. Будущее автоматизированной журналистики. Новый

репортер. Доступно з: http://newreporter.org/2016/01/11/budushhee-

Размещено на Allbest.ru