which does storytelling but integrates Peru Email List social networks from the outset and natively. And concretely, what does that change? Well, quite a bit actually. The heart of our business is working with artificial intelligence solutions. Facebook’s and Google’s algorithms are AI! We know and use these solutions to develop the performance of our customers. But we also have a constant R&D budget in order to discover and use new efficient technologies! ” explains Olivier Méril, Chairman of MV Group. Artificial intelligence, where are we today? “First of all, it is important to know that artificial intelligence is no longer a technique that will exist one day, but a technology that is already in use”
, launches Thomas Léveil, Head of the Development Department, adding “that There it is, we can say that AI is used for real, and not just in labs for tests! ”. As I asked for details, Thomas adds that there are already many examples of AI applications today, especially in the digital rights sector : “YouTube, for example, uses artificial intelligence solutions to immediately detect videos that infringe copyright ”. Better known are the applications of AI in the fields of vision or image recognition : surveillance cameras, autonomous driving… “AI is no longer an abstract concept: its applications are very concrete, and we let’s see more and more of them, ”says Thomas. In the field of marketing , more and more solutions are coming to the market. At MV Group, it is a subject that we know, that we master, which is of capital importance.
But What Exactly Is Ai?
Our businesses are constantly evolving, and our Innovation division allows us to be on constant watch on the subject to identify and test the best solutions. “It should also be noted that AI makes it possible to limit advertising pressure and increase its effectiveness by creating ultra-relevant ads, personalized according to each individual”, adds Nathalie Lucas-Huriau, Director of Innovation and Studies. But what exactly is AI? Although the term artificial intelligence covers several fields, the one that is most often heard is the neural network: “the principle of AI is to simulate in a computer the biological functioning of neurons and their connections. ”,
Explains Thomas. This idea is quite old, since it was in the 1940s that we first heard of artificial neural networks. The idea was then launched in the 1970s / 1980s by teams of researchers, but the computing power of computers was not powerful enough at the time to properly train neural networks. “You have to understand that today, not only do we have the means to collect and store data that did not exist at the time, but in addition, the cost of data storage has collapsed. ”, He continues. It is no longer prohibitive for companies to store huge volumes of data in order to train neural networks.
Ai And Digital Marketing: What Relationship?
Thanks to Cloud Computing solutions , it is also possible to rent computing power from a supplier and only pay for a given time. Companies therefore no longer need to invest in equipment that will have to be profitable. “The drop in the costs of storing and processing data means that today, AI is economically viable”, specifies Thomas, although the most important costs are those of the test phase: “we need a a real tool for creating the neural network, nourishing it and thus training it ”. Because training a neural network is very similar to biology : if we take the example of a baby, it will need a lot of energy and time for its brain to understand how to coordinate its limbs to walk.
But once the walk is acquired, the automatisms are put in place and the energy deployed to ensure the walk is less than that which made it possible to learn it. “One of the great functions of the neural network is the generalization function: concretely, once it is trained, it can recognize a dog in a photo without having to see all the dog photos in the world. ”. Once again, we understood it, like a child who would know how to determine that an animal is a dog among all the dog breeds that exist! “You should still know that this training in generalization requires know-how: