How to determine if an egg is on the stage, a cock inside, or a hen? How do you know what a chicken wants? How to create a hen robot? All these questions are answered in their project Gallus Gallus Roboticus by South African artist and robotics Miranda Moss and American programmer Daniel Brownell. As part of their speech at the festival of science and technology, they talked about how artificial intelligence can be used to raise chickens and how it relates to the problems of humanity.
A robot that sees what is inside the egg
Daniel: My grandfather was a chicken breeder. He skillfully divided the chickens by gender. Oddly enough, this issue was not studied for a long time and caused a large number of problems. One of them, directly related to sex determination, we managed to solve in the last three years with the Gallus Gallus Roboticus project's help.
Miranda: We're trying to build a robot with systems to help look after chickens. The goal of the project is for him to understand whether it is "female" in front of him or "male" while the hen is still in the egg state. From a commercial point of view, this is very important. While working on the project, I certainly took into account my artistic background.
What happens when a niche group gains premium access to technology is being scrutinized specifically. The interaction between robots and humans raises a lot of questions. For example, the military sphere or what happens when we allow groups, sometimes even as absurd as chickens, to interact with robots.
The question arises, how to program and create a robot, if the human activity for him may look like a threat. How do you actually create and develop it? One of the thoughts that comes to the mind of the developers is to put on chickens' costumes so as not to interfere with the robot.
How to combine science and art. Two opinions on this matter
Combining art and science is a kind of holistic idea. Although they seem rather silly to me, there are too many differences between them. We're interested in engineering the project, but it's pretty crazy. The very definition of these disciplines together seems unrealistic and strange. Although we already know the existing working schemes and what art and science are, of course, or STEM in general.
I grew up in South Africa, so I had the opportunity to observe art. This is actually how I grew up, became an artist and how I create art. I also have a bachelor's degree related to art. But I didn’t have access to science education. When I was about 20 years old, I thought it would be cool to do science and start my research. For a long time, I was without any background and technical skills. Where I came from - the study of art included only the study of philosophy and other areas related to this. As a rule, these are the areas in which people develop their soft skills. At the same time, there has traditionally been some division between soft skills and technological skills.
Daniel: Environmental, political issues, and any technology must have a life cycle and consequences. I am studying computer science. I do my development three days a week. I also work at the Cancer Center. As for my business, it focuses more on the environment.
Computer science serves a great purpose. This is the science of problem-solving. Computers have always impressed me because I thought that the computer language is universal, and it seemed that most of the programs inside the computer are not connected with the outside world.
Miranda: But technology is increasingly tied to the lives of people outside. And since computational language is universal, I decided that I needed to do something good for the world. Even though it is as if all good things have already been done, not everyone reaches the stage where they can use their skills to produce something good. I think I've gotten to this stage. Several years ago I decided to break the scientific canon. Everyone around me argued that girls don't make robots because of girls. I wondered why there is such an opinion? I am a girl and I want to create robotics. And I also wanted to understand the extent to which our bodies can act as predictors of the ideas we want to project.
Daniel and I have been thinking in the same direction for a long time and a lot. This concerned the issues of robots and artificial intelligence. For the most part, in the problems with which we are working, we agree and see that no one is ready to deal with many issues that are associated with modern robotics, and with ethical problems. Therefore, we are working on an art project. We can explore specific ideas and specific ideas.
Daniel: Art practice allows you to get out of the rhetoric of the industry. Because the industry wants to solve specific problems, while art is less focused on solving specific problems, it is an art that allows you to come up with something or just have fun.
The first step towards developing robotic chicks
Miranda: Creating a robot is only the first step for our project. And there are already many questions for the next steps. For example, what will happen in several generations? If we already have robotic chickens, how will the chickens interact with them?
Daniel: If we can make a robot that learns to interact with a chicken, it will be amazing. I think I just want a robot to appear that acts meaningfully and does at least something. And as the project develops, maybe we will be able to do something more complicated and add opportunities for the robot to interact with the environment.
Miranda: For us, the question is just assimilation: how to fit robots into the environment. But there’s a weird and fun part that I’m looking forward to working on next year, which is working with chickens. The questions are, why did we decide we know what the chickens want? And how can we program a robot that understands chickens? How can we impose the idea of a sad and cheerful chicken on our robot, so that it can then recognize them? On what basis will this decision be made? I want to understand if we can create a system that will help us understand how to get us as a human species from the same system. Can we train artificial intelligence to speak chicken-like, so that the intelligence recognizes that the chicken is planning to lay an egg at the moment when it plans to lay an egg? And, finally, can we get to the stage in which artificial intelligence will look at the egg and understand that it will be a rooster, and not a chicken, for example.
Daniel: I'm more interested in basic computer vision issues. What are the keys, the intellect will be able to identify in interaction? But I am also interested in the deeper side of this issue. After all, the data that we receive can be used to transform it, create a new simulation, and train the algorithm in the cloud, that is, the data that the robot receives from interaction in the real world can become, in general, the universe of dreams of this robot, in which he will interact over and over again, and test other solutions.
Miranda: Working with engineers, technologists, and programmers, I can more or less speak a technological language, but I struggle to ensure an open mind and a willingness to share what intuition gives. This is difficult to achieve, but we can make it through our interactions. And in this, we were helped by jokes. We watch silly YouTube videos together. These moments are very important because when we watch strange videos, we have brainstormed. When we brainstorm, we generate strange thoughts, and this is the idea.