Kommen Roboter in Den Himmel? / Do Robots Go to Heaven? - De/en/es

Follow us with InsideScience and take a peek over the shoulders of researchers. In this episode: Do robots go to heaven? A journey through the future of robots in society. The year is 2050. As we explore the solar system with complex machines, the first robots enter the daily lives of humankind. A favorite model is the adaptive KIT-O-BOT: "The ideal partner for children who are getting acquainted with the technology of the future through intuitive programming." Lea, a 12 year old girl, got a KIT-O-BOT for her Birthday! KIT-O-BOT quickly becomes her best friend. He helps with her homework, cleans up her room, waters the plants and learns new things from Lea. He brings her to school and picks her up again. But one day Lea waits in vain. Her classmates have long since gone home, but she is still standing alone outside her school and KIT-O-BOT is nowhere to be seen. KIT-O-BOT wakes up with aching sensors on the floor of a long, windowless corridor. -- Ouch, my left diodes! Where am I? As KIT-O-BOT looks around he discovers a strange black cupboard at the end of the corridor that reaches up to the ceiling like a majestic monolith. -- Hey kid, come closer. Fancy a game of chess? -- Chess? No, I don't play that, it has been solved ages ago. Say, are you God? I always imagined heaven to be bigger. -- Heaven? No, heaven is really only for humans. -- Humans, robots, they aren't all that different. Hands, feet, head. Essentially I look like a human. -- Well, humans have built you like that because they always wanted to have a creature that resembles them. Look at me, I do not look even remotely humanoid, but I beat the chess world champion Garry Kasparow in 1997. Those were the days! -- So you are the famous Deep Blue? -- Oh, well! What I meant was that humans have been working on mechanical likenesses since time immemorial. Around 800 B.C., for example, the great poet Homer praised the automata of Hephaistus, the God of Fire and Metalworking. He created two female servants out of golddie that had wheels and could walk to the gathering of gods themselves. -- Hey, I walk like that too, more or less! -- In the year 2 B.C. the Greek physician Galen described the human body using hydraulic principles. About at the same time Heron of Alexandria built animated machines for theatres and temples. This mechanical conception of the human body was long rejected by the Christian church. In the 17th Century however the French physicist and engineer Salomon de Caus summarized numerous automaton plans of his time and raised the question if the human body might work similarly to a clockwork. His English contemporary, Thomas Hobbes pointed out: „All thought is a kind of computation." This led to the great popularity of the android that originated in the 18th century and lasts until today. La Mettrie's Man-Machine, Vaucanson's mechanical duck, E.T.A. Hoffman's Olimpia, Mary Shelley's Frankenstein, Fritz Lang's Metropolis. -- Are those androids real? -- No, they are just a collection of fantasies. The only androids that really exist are humanoid robots, highly complex machines, but mere machines nonetheless. -- But in that case I'm not really in heaven, am I? -- Not at all. You are currently in a kind of sleep mode, but you should already know that. After all, you are a KIT-O-BOT. Didn't your creators supply you with the basic questions of existence and purpose of humanoid robots? -- Maybe, but since the accident I can't really remember much. -- Then allow me a short journey through time into the Year 2012. I would like to introduce you to a few scientists. -- In the development of humanoid robots we often get our inspiration from humans, from how the human body is composed, how the joints are put together, how actuation, motor function and perception work. -- In human form robots can essentially use all the tools of humans. Because they have the same shape we have, they can move well in an environment built for us. This is an ideal combination, I think, especially if robots are to support humans. -- This is important if I want to complete a task together with the robot, like carrying a table from A to B. In that case it is important that the robots can move similarly to us and that we can teach the robot how he is supposed to move. -- On the other hand experience shows that we have to look for alternate solutions in many cases. For instance, an airplane does not flap its wings like a bird. We had to find other methods of keeping a plane in the air. -- The tasks can be completed using other systems. If it is about moving things to and fro in a kitchen environment the form does not need to be humanoid. However this form contributes to the understanding of the human body and also facilitates the interactions of the robot with humans. -- We are used to communicating non-verbally with our fellow humans. Having to communicate explicitly with a robot is tedious and unfamiliar, because we expect differently. If for instance you want to set up a shelf and want to be helped by a robot, you expect that if you reach out your hand you will get exactly what you need. You expect that the robot sees your actions and deduces his next meaningful action to help you. -- If machines are to communicate, i.e. perceive human speech, interpret it and answer to questions, it was discovered that acoustic perception works well already, but only if there isn't much background noise. If too many noises superimpose the speech signal humans apparently compensate for the occurring gaps with their knowledge of the world. We humans have certain expectations about what others would want to say. We assume that what others say makes sense. It is difficult to teach a machine this knowledge in its entirety. -- There are many things in which computers are still far behind human capabilities. This includes the recognition of objects and the analysis of situations. I believe that we still have much work to do before they can reach the abilities of humans. -- In the optimal case we would know how movement control and movement learning works in humans. If we could abstract those theories and models into mathematical equations or develop algorithms that depict the whole process, then we could simply transfer this on to the robot. The difficulty is that we have not fully understood this in humans yet. That makes it very difficult to conduct this transfer. -- Wait a second, if these statements are from 2012 then according to Moore's law I should be able to think at least as well and as fast as a human. -- Well, Moore's law only predicts how fast processors can theoretically become in time, but to be human you still need a consciousness. Look at this. -- I am cautious while speaking of consciousness here. We don't even know yet what consciousness means. We don't even know if our pets, dogs and cats, have a consciousness. I would be cautious in going so far as to expect this in a robot. I think that it wouldn't make sense for our purposes to equip a robot with a consciousness who is supposed to help us in us in our daily lives. -- We already have to weigh the consequences of everything that we do and prepare ourselves for that being possible one day. I think for some areas of application it could indeed be profitable. For menial tasks like loading up the dishwasher a humanoid robot does not need to be conscious to do this. But if it is to be integrated into social etiquette or into a community we will have to think about these things. -- Our generation will not experience technical systems that are actually conscious. For something like that we still lack too much of the understanding of what the consciousness of humans actually is. Understanding the biological model and creating methods or models that can be technologically implemented is still a research goal for the far future. -- But if I don't have free will, then I can't be held responsible for my actions, right? -- I'm glad you asked that. Do you remember what happened this morning? I'll show you. Do you remember this Rubik's Cube? -- Yes, that is my favorite game! -- I can easily believe that. This morning you were crossing the street with this cube in your hand, directly in front of a moving car! -- Oh dear, Lea will be very cross. -- But the worst is yet to come. After the accident the driver did not bring you to the nearest repair shop. No, he simply threw you into a garbage bin. -- But he can't do that! Isn't that a chargeable offense? -- Don't worry, I noted down his license plate. Nonetheless we are going to look for an expert. There ought to be a lawyer behind this door. -- This case is similar to that of a child that kicks a ball onto the street and causes some kind of damage that way, or a cyclist driving onto the road. It would depend, if the child is found to be responsible. Because children aren't responsible the parents would have to take the responsibility for the use of the robot by the children. So far this is nothing new, we can just run the judicial process, as we already know it. But concerning the other thing: Because the robot is not autonomous the driver has damaged a piece of property. Robots are property just like everything else. That means that a driver who is involved in an accident and who doesn't stop but leaves the scene of the accident is committing a hit-and-run offence. Another thing... It gets more interesting when robots are autonomous one day and can say "I". But I would bet that when the first robot comes and says "I", people will say it was programmed to say "I" and that it doesn't really have a sense of self. It goes on of course, what will we do if we destroy it and it cries out in pain? Here we come back to the old Turing test: the question what happens when I can no longer determine if my communication partner is a human or a machine. Then it becomes difficult because there is no third party observer that can look inside the robot and say that you may only see a robot from the outside but actually it has feelings like a human. That is not possible. -- But am I responsible for the accident, or not? -- Difficult question. Maybe we should think some more about the questions that arise from humans and robots living together. In the past the cohabitation of humans and humanoid robots was neither technologically nor culturally seen as natural. In the year 2012 there already were robots that could perform certain tasks, but none of them were advanced enough to live with a human or even to perform the simplest tasks that are easy for you now. -- Ah, you mean like doing the dishes and folding up clothes and hey, look, he can do the Rubik's Cube much faster than I! Wow, I love that game. -- Back then humans only knew robots from science fiction, but that delivered a rather skewed picture in some regards. -- Science fiction or generally the literary, cinematic or other depictions of robots only deliver a wrong picture if we expected the actual technological stage of development to be shown. They show technological futures rather than the technological present and they deliver these futures with very strong judgment. They are either utopistically positive, what could be possible, or dystopistically negative, what kind of bad could come from this if the side effects disregarded by the designers come to bear. -- Robots in movies are generally conceptualized as helpers for humans. The Golem for instance was supposed to protect the Jewish community from attacks or as seen in the case of the false Maria in "Metropolis", who doesn't have a name by the way, who is supposed to replace a lost lover. Or the famous Gort in "The Day the Earth stood still", who is actually a universal peacekeeper who takes away the human's weapons to enforce peace. But of course in order to get an interesting movie something needs to go wrong. Otherwise there is no drama. When the robot malfunctions everything gets interesting. It is fascinating to watch how in the cases where the robot doesn't work as intended it is often the human factor that becomes important. -- There are people who use the robots to pursue their own selfish interests. In this case we often find that the nature of the robot is obscured, so that the robot presents itself as human. In this lies the most evil and perfidious aspect, that we have a being that can be controlled from the outside but that appears to other humans to be a real human that manipulates them, leads them and instigates them to a variety of deeds. This is the case in "Metropolis". -- The movies about robots ask, I believe, the serious question of the similarities between humans and machines. We have this beautiful scene in "Blade Runner" where Roy Betty, the leader of the replicants is sitting on the roof in the rain. He is telling his enemy, who had just tried to kill him but then saved his life, his last memories, shortly before he dies because his internal life time has run out. Roy Betty is set up so that he will die at a certain time. He feels this moment. Then he shares his last unique memories with his opponent. That is a wonderful scene. -- I think that these fictional depictions of robots strongly influence the concept of robots in the public mind. Probably much more so than the real robots, which exist, because these are a lot less spectacular. -- Many people back then could not understand how far research had progressed. No one could imagine that humans and robots would live together one day. -- I would sooner talk to a screen at a cashpoint, or anywhere else where there is an input prompt than with a faux human, a robot that sits in front of me in the guise of a human and interacts with me would most likely alienate me. -- "Humanoid" means "similar to human" and instantly evokes associations and provides a blank slate for the projection of longings, wishes and needs one also has towards humans. Interestingly qualitative studies that I know of in regard to humanoid robots in the care field have shown that people decidedly do not want humanoid robots in their surroundings. They don't mind having technical support or technical assistance systems, as they are often called, but they shouldn't be humanoid. Probably because it leads to confusion. It wouldn't be clear which category a humanoid robot occupies in my life, in my daily routine or in my surroundings. -- Take for instance the care of the elderly. I don't believe that a robot should care for an elderly person. The daily tasks of the nurses that do not involve humans should however be done by robots, for instance fetch and carry tasks or surveillance. The actual human contact should still be conducted by the nurses or the family members. My hope is that we can delegate all the menial tasks to robots so that we have more time to spend with our fellow humans. -- That is strange. For Lea this is completely normal. -- I'm not saying that all humans were afraid of robots There were always people who were fascinated by the idea. Children too could always be enthusiastic about robots. It seems to be human nature. -- I think the interesting thing for students is to have a device they can control and that they can teach things. That is the interesting thing about it. They get a problem to solve and they have to think ahead what the robot should do and how it can be taught. -- I find it fascinating that I can grip things with a mechanical arm. -- I'm also impressed that they can see colors and grasp things. -- I'm just really fascinated by how these things work. How the robot is built or how the robot does what I tell it to is not quite my thing. But how I tell it things, that's what mainly interests me. -- I'm more interested in what it can do. It is interesting to see what a robot can achieve if he is programmed correctly. -- They learn through success and failure. They also learn that their thinking works differently than it does for a robot. -- Here we have the NXT and it should now complete a task. It's supposed to drive forward, press the lever so that the syringe rolls into the field. We simply have to practice programming all that -- There are many elementary pupils who have come in contact with Lego Mindstorms or similar products. I believe that it is more likely the older generation that got in contact with robots over science fiction, like me for instance. Of the top of my head I could name a dear children's book of mine, "Robbi, Tobbi and the Fliewatüüt". It tells of a robot that has to pass its robot test and that enlists the help of Tobbi in order to complete the test. They go on many adventures together. -- Oh, I miss Lea! I wish I could go back to her. -- It's still too early for that. I have to show you another thing: The robots of the future and how they will change the world and human perception. -- We can adapt to systems that are versatile, that complete tasks in private environments as well as in public ones and in factory production. That means systems that fetch and carry in our homes and that can watch the house while we're gone, tidy the child's room or clean the bathroom. The visions go further. Based on these systems robots could be developed that are used as additive components for people with physical or sensory disabilities that make life easier or allow people to live for longer in their own homes. -- I believe that the development of the robot mechanics will continue rapidly, which means we will see robots that see, jump, run or crawl. We will see robots that can play soccer and many other things. In my opinion, the mental abilities will develop far slower. This supports the approach of trying to implant the abilities of humans into the robots, meaning remote control. In such a situation like in Fukushima it would be possible to send a robot that has the mechanical skills, but in principle itself cannot do much, then a person would teleoperate the robot. -- I imagine that it won't be like what we know from science fiction movies, that we have a humanoid robot that can complete every task, but that we will have different solutions for different tasks. For instance a robot that is responsible for hoovering. Something like that would of course be built differently than, for example, a robot for helping people with the packing and unpacking of groceries. -- The robots in the movies will always be ahead because they look far more spectacular than in reality. Real robots are designed in accordance with their very specialized functionality. On the other hand we have an interesting trend in the 'cyborgisation' of humans meaning that the human body can be augmented with mechanical or electronic parts. That would be another prospect. Furthermore science also has a strong wish for the comparison between human and machine. Scientists are trying to create the most humanoid robot possible. Experimentation in that field is sure to continue. -- In the last five to ten years technology has reached into our daily lives. That means that we can expect service robots and service technology to also catch on. I believe we must keep in mind what we want to protect, what we don't want to lose. This includes categories of environment and time, leisure and above all else the understanding of the high quality that human communicative or interactive environments provide and what skills they offer. -- If we simply assume that robots are no more than tools then robots are going to be new tools that help us solve problems that we could not solve before. If looked at from this point of view we can indeed look into the future optimistically. But at the same time we mustn't lose sight of the risks. That is what legislation is for. With the means of legislation we try to control those risks. We cannot predict or control everything because we cannot conceive of every possible scenario. That means that we will always have to decide what economics call a decision under uncertainty. That is something we just have to live with. -- I want to be part of this future, too! KIT-O-BOT wakes up screaming on the garbage dump. Daylight has given his photon-sensitive skin an energy-bath and now his storage batteries are full again and his operating system has booted up. In about an hour KIT-O-BOT is back home. Despite the fact that he smells a little bad, Lea is very happy, and KIT-O-BOT is too, in so far as robots can be happy. As Lea hugs him, something moves inside KIT-O-BOT. Lea opens the small compartment in the robots belly and finds the Rubik's Cube. She hands it to KIT-O-BOT inquiringly. He solves the Rubik's Cube and hands it to her. On the side turned towards her Lea reads E N D E 4 2. Is that a licence plate number?