The first specialized meeting of ielligence ambassadors with the topic The role of artificial ielligence in the media and news industryIt was held on Thursday 24 Azar.
In this meeting, which was held at Ilsa Creative and Innovation House, guests from Arman Rayan Sharif Company, Ati Project Data Processing Company and Sports 3 Media hosted by Ecomotive were prese and discussed the topic.
The discussion started with Saleh Esadi from Arman Rayan Sharif Company and he gave explanations regarding the activities and projects carried out by this company in the field of media. The launch of the first cloud platform of artificial ielligence, which was unveiled by Dr. Sattari, and so far 4 million sounds, videos and images have been uploaded to it, is one of the latest projects carried out by Arman Rayan Sharif and the text-to-speech conversion service Smart), video-in-video or audio-in-video search services, as well as the news recommender service, are among other projects that have been carried out by this company according to the needs of the media industry and the capacity of the collection.
In the following, Mr. Arash Mousavi, Chairman of the Board of Directors of Ati Project Data Processing Company, regarding the poteial of artificial ielligence in Iran to play a role in the media field, stated: First of all, it must be accepted that today the term artificial ielligence has become a widely used term; As it happened to the word nano in the past years. In this situation, it can be said that many times the use of artificial ielligence technology has become a luxury and sometimes there is no economic justification for practical action in this field. You will notice that a serious and importa need is not being answered. This is why our collection and our similar collections have gone io this story much less, but on the other hand, we slowly felt a serious need in some places and found some input.
We have not done much work in the field of media and artificial ielligence. Over the past year we’ve made a very small start on ideifying what commes should and shouldn’t be published. Also, setting up a helper for SEO where you can, for example, tag commes or categorize commes has been one of the other efforts of our group in this field, which, of course, has not been used seriously.
One of the issues that has preveed us from seriously eering this field is privacy concerns and various legal issues. We deal with a lot of information and to use it and provide it to differe services, we have always had concerns about whether the information is given or not. This happens even in cases such as news agency sites that are used by the public. All in all, there is such a complex story, which is not unsolvable, but it is something that we are facing, and these have caused us not to do much serious work in this sense.
Naturally, I think that this situation is not only for our collection and unfortunately, the atmosphere of this field is like this.
Mousavi further said: There is no doubt about the fact that artificial ielligence has many capacities, but sometimes the expectations of this technology are high, and we forget that a little patience is also required for the proper use of this technology.
For example, let’s check one of the artificial ielligence services, chatbots. When you wa to add an operator in your system, first you don’t have to put the operator instead of the curre reference operator. An operator is first an assista to another operator and even later; That is, he first understands what the problem is and then gradually becomes an assista, operator and chief operator. Ielligence must follow the same path. This is an unreasonable expectation where we say that artificial ielligence is going to be our main operator from tomorrow.
Alireza Anbarlui, product manager of Sport 3 media, said the situation of Iranian media, especially Sport 3, in the use of artificial ielligence technology and the challenges of the media space in this regard:
Regarding the comme refineme system, we are doing a lot of work in Sports Three. It started almost two years ago, and our most terrible challenge was that we had to manually refine our comme system, which happened very late and users from They were unhappy about this. A series of very simple reviews were done in marketing and we came to the conclusion that we should start a revolution in the discussion of refining our opinions, and our requireme was that, first of all, there is a system of refining opinions that rejects an opinion that is very specific and that The blacklist system simply did not answer us.
In this regard, we talked with various groups and finally we implemeed a model that was able to automatically reject up to 30% of the commes that had problems, and this model is still being updated, but we still have many challenges, our challenge is this. that we have differe ethnicities in Iran and each of them has a special dialect and terms, and this makes our work very difficult in the field of opinion filtering, and because the updating of this filtering system is usually done mohly, we need this We wa to coinue to have the manual filtering system because the behavior model of the audience and the methods used by them sometimes change daily. Therefore, increasing the speed of updating the opinion filtering system is one of our importa needs in the media.
Also, in our collection, we do other things in the topic of image processing, which is now mostly in the topic of the aenna collection, and the summary of the game is that we eer the fields captured by the aenna io an artificial ielligence system. This system detects critical game scenes and merges them together and loads game summaries in a fraction of a second.
At the end of his speech, he said: “In Iran, we are in a relatively good situation in the field of artificial ielligence, and the fact that groups are thinking about this technology and that many elite young people are working in this field is really very encouraging.” Despite the lack of access to modern education and the existence of many challenges, we will definitely be in a better position in this field in a few years.
Mr. Safaei from Data Parzahan Ati project, in line with the challenges of using ielligence in the media, said:
One problem is that there are a lot of people working on ielligence, but none of them have come to open up the data they’ve come across or the algorithms they’ve partially developed, and allow others to add something to or from it. reduce A series of data that is abunda in English and well taught also exists in Farsi, but it is so little and everyone has worked separately that it is difficult to access them. In fact, companies turn recorded data io their iernal value.
How do you rate this article?
