The incoming AI revolution in media
From fighting fake news to monetizing old video content and bringing global audiences to local events, artificial intelligence could completely transform the way media outlets operate and make money. This was the subject of a fascinating conversation at Maddykeynote 2021.
Whether it plays go, drives cars or learns to speak, artificial intelligence (AI) is now everywhere, benefitting a wide range of industries. Like many others, the media industry has already been radically changed by the rise of the internet and new technology. It is thus all the more logical to reflect on what’s coming next, and there’s no doubt that artificial intelligence will play a significant role.
I was at MaddyKeynote 2021 to discuss this topic with two professionals of the industry, Guillaume Doret, CEO of Synchronized, and Arnaud Maupin, director of innovation at Group TF1.
Monetizing old news bulletins
According to Guillaume Doret, AI provides cable news channels with a tremendous opportunity to index and monetize their video archives. “For now, a video is just a file with a name. When you open it, you only see encrypted data, which is a missed opportunity, because it stops you from navigating inside the file and extracting relevant information. As an example, while the broadcast news from April 15, 2019 may be outdated and irrelevant, it also happens to be the day that Notre-Dame burned, which can still prove valuable today.
Currently, you can manually index the video, but it is cumbersome and costly, hence not a scalable process. With artificial intelligence, you can automatically search the file, find the sequence you’re looking for and monetize it. That’s the kind of solutions we provide at Synchronized,” he explained.
Not only can artificial intelligence search the video for specific information, it can also put this information in the right shape. “We’re currently working on a technology to automatically generate a trailer. That’s a big challenge for streaming platforms, who sometimes have a library of 10,000 movies. Starting from the whole movie, artificial intelligence can do reverse engineering, and come up instantly with a 30/second version of it. You then have a trailer that would have otherwise taken at least two days and a lot of money to produce, which is very useful when you’re dealing with huge volumes of data,” said Guillaume Doret.
Spotting fake news
Another challenge that the media industry is currently facing, of course, is fake news.The Internet and new technology gave everyone a voice. We are all able to broadcast live videos, for example, which is a tremendous opportunity but also a challenge, making it hard to know what is good information and what isn’t. Artificial intelligence can help by ensuring that a piece of information comes from a reliable source. As an example, you can know who the person broadcasting is and what he did before. It can also help you see the whole journey that the video you’re watching went through, from beginning to end, thus knowing whether there was forgery, manipulation or not.
Now let’s say you’re watching a debate between two politicians arguing about a specific statistic. Both insist they’re right. Which one is telling the truth? This can prove quite frustrating. Artificial intelligence can provide some background information. “At TF1, we’re exploring how artificial intelligence can do some fact checking during a live debate. In the future, in the same way that Word underlines a word in red when the orthography isn’t correct, you’ll be able to see the piece of information at the bottom on the screen, in green if it’s true, and red if it’s not,” explained Arnaud Maupin, director of innovation at Group TF1.
“With the possibility of increasing personalization through AI, targeting a specific content to a specific audience, plus technologies such as virtual and augmented reality, you start to see a media landscape that will be dramatically different.”
Finally, artificial intelligence provides media companies with scalability. During the Olympic and Paralympic Games Tokyo 2020, more than 100 million live events were securely broadcasted through different media around the world. You absolutely need artificial intelligence to deal with that kind of volume and distribute content in a reliable way.
As 5G becomes mainstream, edge computing will help media companies go even further when it comes to scalability. Nowadays, 80% of data is processed in data centers and 20% on the edge. In the future, it will be the opposite. It means that to capture a live event, you’ll be able to use nodes of computers working on the edge, dramatically reducing the amount of costly and cumbersome material you now need to be able to broadcast.
This will lead to an explosion in the number of live events being shared, and will allow some small, regional events to broadcast to the whole world, making much more money in the process. Add the possibility of increasing personalization through artificial intelligence, targeting a specific content to a specific audience, plus technologies such as virtual and augmented reality, and you start to see a media landscape that will be dramatically different.
A significant challenge that media companies will face in the years to come is the key role that big American tech companies are already playing when it comes to regulating information. We’ve recently seen, for example, how Facebook deactivated the accounts of some individuals in Afghanistan so that the Taliban couldn’t target them. These big tech companies are becoming central to spreading and regulating information. Is it truly their role? I don’t know, but that’s a discussion we’ll need to have for sure.