"Helicopter" is a conceptual visual project exploring the potential of artificial intelligence in creating architectural presentations. It is based on the idea of ​​how a limited set of source images can be used to create a coherent story about the birth and development of a residential complex. The project is assembled from just three source images, which were interpreted and enhanced using neural networks. The result is a series of images demonstrating the various stages of construction—from excavation to the formation of the architectural volume. The flexibility of this approach is particularly noteworthy: it allows for virtually any person to be "incorporated" into a project—for example, the CEO of a construction company—and their image to be tailored to the task. Using neural networks, clothing can be modified, accessories added, and vehicles (cars, helicopters) can be added, creating the desired visual context without complex filming. However, it's important to consider the technology's limitations: inaccurate rendering of architecture and the surrounding environment is possible. The image will appear realistic overall, but the neural network can add its own elements or simplify details.