Radar Armenia‘s interlocutor is IT and media consultant Artur Papyan.
– How do you assess the role of artificial intelligence in the media field: do they help or hinder?
– Today, in the media field, when there is an acute need and shortage of resources, artificial intelligence, which can edit texts very well, accelerate editing, image generation, and many editing functions, acts as a lifeline for the media because it allows you to significantly reduce the costs of content production, processing, and even distribution tools. It is possible to achieve such a result with 1 to 2 journalists, one editor, and one artificial intelligence specialist, which could have been completed years ago with much larger teams.
– What problems do you see in journalism related to artificial intelligence? What challenges are there?
– Regarding challenges, the problem is that journalists must be transparent about what artificial intelligence tools and technologies they use. In this regard, the Armenian Media Self-Regulation Initiative has already developed an ethical code of conduct for using artificial intelligence, which states that journalists must be transparent.
For example, if a photo was generated by artificial intelligence, this should be noted; if it was created for video-illustration purposes, this should also be noted. In the case of data processing, it is also desirable to note that they were processed using artificial intelligence because it is understandable that artificial intelligence has prejudices arising from the data it was created with. Let’s assume that all the content was made in the United States, based on American technologies and concepts. Then you ask about Armenia and artificial intelligence answers specifically from its training material. It can make decisions incompatible with Armenian reality. Or, if you give artificial intelligence 10 thousand lines of data to process, it will proceed from or take as a basis the Training Data, the model with which it was trained. This can lead to specific problems; its Training Data Set may bias the results, which should be realized.
The second key provision in the Artificial Intelligence Code of Conduct of the Media Self-Regulation Initiative states that artificial intelligence developments must be checked, rechecked, verified, and never relied on solely.
– A series of programs has begun: ChatGPT, DeepSeek, Qwen, etc. What will this lead to?
The race for different programs is currently beneficial to consumers. At first, this technology was “behind the fence”; it was paid, and ChatGPT was trying to generate money. Now, it is becoming a standard product, accessible to everyone. The price will decrease, and it will become an everyday thing everyone uses. We will actively envision our lives as inseparable from artificial intelligence, just as today it is difficult to imagine life without a mobile phone.
