How artificial intelligence is changing modern television and the film industry.

Artificial intelligence is getting closer to us thanks to digital media and the film industry. While we have no idea how self-learning systems can help produce more quality content in less time, but soon it will be as familiar as using a smartphone.

When will artificial intelligence (AI) completely replace creative people? This will probably not happen on a global scale, but the first attempts to increase creative productivity have been successful. This mainly concerns simplification of routine actions - user interaction with the TV, selection of actors, preparation of news. But there are also more impressive examples - for example, a full-fledged feature film was assembled using AI!

Learn more about Interactive TV from OnlineLime

We are so used to Smart TVs  in modern TVs that we hardly expect anything qualitatively new from this feature. In fact, it has great potential for development. The latest CES showed that AI will soon become an important part of Smart TV and offer users a new way to interact with TV.

Do you often lose your remote control? This will not happen again, because the AI ​​will process voice commands on the fly. You will be able to turn on the channel, start the game console, find any information about the objects that are displayed on the screen. Imagine that the Siri counterpart is now built into your Smart TV. The voice assistant will not only turn on a certain channel, but will find it on its own by the name of a movie or TV show. Watching a movie? Ask any question about an actor, director, or even the city in which the events unfold. While watching a sports broadcast, learn more about a particular player: his biography, career history, game statistics. During a show or movie about, for example, New York, ask your virtual assistant about the weather or traffic jams in the city. And all this is in direct reach,

Naturally, you can change the volume of the sound with your voice, and adjust the brightness of the picture, find out the rating of the film or get a ready-made selection of interesting YouTube videos based on your preferences. Improving the quality of the original image is not a problem. The latter is exactly what Samsung is planning: it is working on tools that transform 4K content into full 8K.

Artificial intelligence in Smart TV can be easily integrated with the smart home. When all systems are in sync, the TV will automatically dim the lights to create a comfortable environment for watching a movie or TV show at night. And he will also launch a sports YouTube channel after turning on the bike path.

How AI selects TV and film actors

Most likely, you paid attention to the credits of the film, where not only the actors are indicated, but also the person who is responsible for their selection. In a few years, this position may disappear - casting specialists will be replaced by artificial intelligence.

Can you imagine how many actors are applying for roles in one film? Thousands! From everyone you need to get a resume, photos, videos, audition, fill out documents, agree on working conditions. All these are huge time and human resources.

In 2013, the EasyCast platform appeared, which automates most of these processes, and communication with actors no longer requires such a time investment. Here, artificial intelligence performs part of the creative functions of a casting director and selects suitable candidates using neural networks. Over the past five years, the system has been continuously trained and improved, and now, in a single company, 80% of the actors are selected by it, and not living people.

The first experience of using AI in casting was classified - the producers of the film did not know that the casting would be done not by a living person, but by a machine.

How it works: The project director or his assistant fills out an online form and starts a search in a large database with one button. The system searches until it finds a sufficient number of matches. In some cases, only a textual description of the actor's image is used (faces are compared for similarity), in others, there are many auxiliary criteria that ultimately help to get more accurate results.

Now such systems select not only movie actors, but also participants and viewers of popular TV shows. Every month, there are tens of thousands of questionnaires that previously would have had to be processed manually. By the way, the automation of the process significantly increases the chances of getting into the movies and on television thanks to professional and external data, and not “by acquaintance”. Nothing has changed only for the main and supporting roles - they still operate with traditional search schemes.

Another interesting area is when artificial intelligence analyzes scenarios. A special algorithm with a high probability determines whether the film will "shoot" or fail at the box office.

The first movie edited by artificial intelligence

Film buffs will be disappointed, but the next ten years will see a revolution in the film industry. Netflix recently picked up the archives for Orson Welles' The Other Side of the Wind, which began filming back in the 70s. The project was never completed, although the necessary materials for this were available. Then the specialists of Netflix put them into a system built on the basis of artificial intelligence, and that significantly improved the quality of the picture. Viewers received a full-fledged 4K picture with excellent resolution at the output, and the company saved on human resources. Forty years ago, Orson Welles would have had to cut and then glue a film from a huge number of fragments, and AI replaced routine with software processing to produce an even better image.

Top 5 Artificial Intelligence Movies of the Decade:

• WALL-E (animation WALL-E, 2008);
• She (feature film Her, 2013);
• Lucy (feature film Lucy, 2014);
• Out of the Machine (feature film ExMachina, 2015);
• "Westworld" (TV series, Westworld, 2016).

How AI helps create news

The media is a continuously working machine that digests gigabytes of information every day. It is necessary to find, select, process and correctly present it to the target audience - viewers. If at the level of a city TV channel a small team can cope with this, then on a national scale it is difficult to do.

News in general can hardly be called “creativity” - this is exactly the direction where it is easy to derive a formula or check how the material meets the criteria. Therefore, artificial intelligence organically fits into the creation of news content. In the US, more than two-thirds of the media are already using self-learning systems. They check the accuracy of facts, select tags (it is much easier to find the information you need in a structured database!), Use marketing tools, and more.

Artificial intelligence in news media helps both TV viewers and journalists. The former are less likely to burn out at work, performing a large amount of routine and not at all creative work. The latter receive more quality news - they are not particularly interested in what remains behind the scenes.

Urbs Media service for regional media

The British Press Association initiated the creation of Urbs Media service for urban and regional media. This platform regularly hosts hundreds of semi-finished stories that local sites and offline publications use for their own purposes. In addition, journalists have access to a huge database of statistics - about politics, medicine, society, etc. They can be used to create short local notes, which, again, are distributed in local publications.

Dreamwriter program for creating reports on forums and conferences

Dreamwriter's automated system, which also uses the power of artificial intelligence, has a narrower specialization. It is designed to process content from forums, conferences, seminars, public meetings and other public events. Dreamwriter processes live speeches of speakers and converts them into text form, and journalists use it for their own purposes. Imagine how many hours you can save on transcribing speeches if the system processes several thousand stories a day!

Automated processing of interviews at the BBC

The BBC service fundamentally does not publish news generated by artificial intelligence (although it is a matter of time), but uses other systems - for example, to automate the processing of interviews. Another interesting feature is finding trends based on open source analysis. News packets from separate fragments are assembled in a similar way. From several dozen pieces, the system adds a decent quality news release, where each fragment fits organically into the overall "picture".

Artificial intelligence is unlikely to replace the news editor, but it will help reduce the staff of journalists and specialists in related fields, who were required to do low-skilled work. Try to calculate how much an average employee will spend on reviewing tens of thousands of archive fragments and assigning tags to them! Of course, the final decision is made by a person - he chooses fragments that are interesting to viewers. But it is the machine that offers these pieces of video, to which the editor sets specific search parameters. Each time artificial intelligence learns and becomes "smarter" - which means that it selects better quality fragments further. He will be able to hone his skills endlessly!

AI-powered sporting events

Ever noticed how cameras work during important sporting events? Unlikely: you were so passionate about the game that you did not see the obvious - how "smart" systems create a picture on your screen that will take your breath away. This artificial intelligence tracks the progress of the game, analyzes the most stressful moments and focuses exactly where something interesting happens. He also creates replays, zooms in for more detail and inserts ads where they will attract more TV viewers' attention.

This is not a complete list of the possibilities that AI opens up for TV people. Would you like to see “smart TV” available to everyone?