Artistic AI "Vincent" turns your sketches into masterpieces by Van Gogh

Cambridge scientists, together with colleagues from the Prisma studio, announced the creation of a new type of artificial intelligence. He was nicknamed "Vincent" for his penchant for expression and initiative - the system does not obediently fulfill the task of a person, it is designed in such a way as to strive to draw on its own. This is one of the first-borns of the concept of multi-AI, a technology unthinkable for yesterday.

The developers of Vincent have combined two machine learning techniques in one system. The "technical" half is responsible for image processing - the AI ​​has fed countless examples of painting and painting techniques, from the Renaissance to Picasso. As a result, Vincent's robotic stylus can draw in any known style. The “emotional” part is based on advanced experimental technologies and analyzes the artist's actions on the fly, tries to understand his intentions and turn a simple first line into a finished drawing.

It is useless for "Vincent" to give specific commands, he acts independently and in parallel with the human artist. AI seems to be trying with all its might to create, but still cannot throw off the chains of limitations - it needs to be given at least a few lines, a basis, so that a picture is created on its basis. But it is extremely difficult to predict which one exactly, since the “way of thinking” of “Vincent” is very reminiscent of the whims of unrecognized talents.

The combination of two machine learning systems that serve different areas of work to obtain a single result is multi-AI, in a simplified form. For example, a robotic chauffeur equipped with such an AI will be able to combine the transportation of fragile goods with an independent analysis of traffic on the road. The complex, controversial, but interesting future of artificial intelligence.