Can a Robot Paint Like a Human? Exploring the Future of Art
My name is Vladimir Tsimberg, and I am the founder of Robohood Inc. We created a groundbreaking algorithm that enables an AI-driven robotic arm to transfer generated or original digital images onto canvas, creating its own unique artistic style.
This technology empowers individuals without artistic skills to explore their creativity. For professional artists, it offers a novel tool to push the boundaries of their artistry, from engaging in interactive performances with machines to achieving new heights in painting quality. I have witnessed this transformative potential firsthand.
Malcolm Gladwell, along with researchers from Erickson & Co., introduced the 10,000-hour rule, suggesting that mastering any skill requires at least 10,000 hours of dedicated practice. This principle underscores the classic correlation between effort and achievement.
From an early age, I was deeply passionate about painting, experimenting with various colors and textures. My relentless pursuit of artistic improvement through studying relevant literature and constant practice honed my skills, even though I never became a professional artist. This experience sparked the idea of transforming those countless hours into technological innovation.
Books, articles, fine arts, and musical compositions encapsulate accumulated knowledge and experience that can be shared. Yet, the challenge remains: how can we transfer skills and abilities? Many experts worry that their expertise cannot be passed on, and the opportunities to monetize this knowledge are limited. Thankfully, technology has provided us with a means to capitalize on intellectual property.
Take endocrinologists, for instance. Traditionally, they provided their expertise through direct interactions with patients. As technology advanced, physicians began embedding their knowledge in devices such as insulin pumps, which monitor blood glucose levels and deliver insulin doses as needed. This innovation has reduced the frequency of doctor visits. However, endocrinologists are still essential to monitor the proper functioning of these systems and provide guidance. But the core knowledge has been successfully integrated into the technology.
A similar transformation took place in photography. Initially, it lagged in quality and accessibility compared to the artistry of photorealistic portraits. Taking a photograph was once a laborious and expensive process. Today, thanks to smartphones, almost everyone has a camera in their pocket. Yet artists continue to create paintings, and there’s no sense that one medium is replacing the other.
Technological advances should not be judged solely on their current state. I remember taking photos with a 0.3-megapixel camera phone, while today many people have abandoned professional cameras because smartphone photos are of comparable quality.
Similarly, our Robotic artist has faced criticism for its rough brush strokes, challenges with color blending, and difficulties in painting eyes. However, our project is on the classic trajectory of early technological development. We recognize that there is considerable room for improvement. As Allen Dulles aptly noted, “You can fool the people who are only interested in facts, but you can't fool those who are interested in trends.”
The ongoing industrial revolution, often referred to as Industry 4.0, is characterized by the integration of robotics into various processes, including painting. This trend is a testament to the transformative power of artificial intelligence and the pervasive influence of information technology across all sectors.
Critics may point out certain ‘flaws’ in robotic art, but these are actually distinctive features. Our Robotic artists emulate the style of the Impressionists, achieved through deliberate, accentuated brush strokes, delicate gradients, and seamless color blending directly on the canvas. This technique mirrors the approach of Vincent van Gogh, and it’s why our robotic portraits often elicit such strong emotional responses from viewers.
Our robots can produce 1,000 brush strokes per hour, completing anything from detailed portraits to abstract works within 7–8 hours, depending on complexity. Artistic styles are executed using algorithms designed for acrylic and oil painting, incorporating various artistic effects and brush sizes. For quicker projects, the robot artist can create sketches in as little as thirty minutes. Our R&D department has increased the speed of artwork creation by about 30% in the last six months.
The journey of technological improvement is challenging, but the results are proudly displayed in our Robohood art studios. Whether the works are original creations by artists or generated by neural networks, they are primarily produced on 40x40 cm canvases, the maximum size our robots can handle. However, larger paintings are also possible, and small canvases can be freely combined into large compositions.
In this process, the quality of the painting improves significantly with the number of brush strokes, making the details on the canvas more expressive.
Our team is currently focused on developing advanced color algorithms that go beyond digital color mixing to explore more intricate methods of blending physical paint. Our developers work closely with consulting artists to understand the nuances of different paints, brush shapes, and sizes.
It’s like an art school for robots, with our development team serving as instructors, translating all of their expertise and skills into code and technology that can be shared. This is the essence of knowledge transfer. One of our primary objectives is to create technology that can function autonomously and independently across various fields, including education, fashion, medicine, and event planning. The capabilities of our Robotic artist have already gained recognition in these areas.
At Robohood, we believe in the power of collaboration. Our team comprises individuals who may not know each other personally, but work together to achieve remarkable results. We use open source Python libraries for our software development and are building a knowledge cloud around our project. This knowledge cloud follows the open source philosophy, allowing anyone who joins our team to contribute their expertise, fostering a collaborative platform.
This platformization allows participants to interact at a higher level, working together in a system where different actors engage in a market-oriented environment. Specialists in graphics and painting — artists, critics, and collectors — can join our team to help develop algorithms. This collaborative approach not only reduces the cost of the final product but also enhances customization options, such as the details of eyes, brushes, and hands in the artwork.
We can draw inspiration from history by emulating the collaborative practices of the past. During the Renaissance, masters often worked together to create masterpieces. The platformization of Rubens’ workshop exemplifies this, as he invited skilled artists like Jan Brueghel, Frans Snyders, and Anthony van Dyck to contribute their expertise in specific areas, such as animals and still life, to different parts of a painting.
Similarly, Titian and Giorgione collaborated on different parts of a painting, while Cornelis van Poelenburch enlisted landscape painters like Bartholomeus van Bassen and Jan Both to create backgrounds to which he later added figures.
Robohood is at the beginning of its journey, with developers currently experimenting with machine vision. For example, we are challenging our robot with tasks similar to those faced by human artists, such as painting from memory after being shown a photograph. This is a challenging task, even for professionals. As our robot undergoes this “blindfold” test, its work quality will continue to improve. This concept has potential applications beyond robotic painting.
Ultimately, in addressing the question posed, it is not about comparing a robot to a true master. Many artists acknowledge that this tool is advanced enough to enhance their artistic expression, refine their technique, and serve as a robotic brush or collaborative instrument. It is unlikely that Robotic artists will seek to surpass humans; instead, they excel as collaborative partners in the creative process.
