MOSCOW, April 21In the near future, neural networks will learn to assess the amount of damage from accidents and help at metallurgical plants, while people will not stop using them to generate music and images, said Konstantin Romanov, director of artificial intelligence and digital products at Beeline.
«In the near future, those application scenarios that are now considered innovative will be developed on the Russian market. This, for example, is a service for assessing car damage using a neural network, which is able to assess the scale of damage to a car body after an accident from an image, to detect traces of dirt, snow on the surface of the car , stickers, etc. An obvious application scenario is an objective assessment for insurance payments for damages,» he said.
According to him, now one can observe the prerequisites for the emergence and penetration into life of strong or general artificial intelligence (AGI) — these are, for example, chat bots and other generative systems that are used in various businesses. In addition, neural networks will continue to be used to create musical and text passages, as well as images.
At the same time, another atypical development scenario is the widespread use of computer vision technology in calculating the volume of roundwood. Romanov clarified that today, for example, one of the timber holdings in Russia is using this method. And in the future, artificial intelligence will increasingly be used to «detect steel in the slag stream at metallurgical plants.»
However, despite this, the expert predicts that in the next 10-15 years, decisions will still be made by a person, albeit based on the conclusions and analytics that a trained model can collect.