The title of this talk might seem a little odd, to even use the phrase machine life might seem a little presumptuous. Can machines actually be alive? This is a conversation that people have had for a long time, and it's a theme that's going to play into a lot of our discussions about how to model the human mind or living minds through the use of machines, primarily computers. When I say this is an old conversation, the ancients actually wrote and talked about the allure of machines that could behave in life like ways. In the Iliad by Homer, there is a description for example of the god of technology, Hephaestus, who has at his beck and call a group of machines. They're described as golden maidens who are able to help in building and constructing the shield of Achilles. The golden maidens are described in terms that today we would think of as kind of similar to robots, their machines that move and that imitate human beings. It's not clear at all what kind of machines Homer might have been familiar, with the Iliad was composed some time around maybe 800 BC. It's not clear what kind of machines he might have been familiar with, but he was able to imagine that machines could be constructed that would imitate life. One of the earliest actual examples that we have knowledge of, is the machine shown on this slide here by Hero of Alexandria who lived in the first century CE, and who constructed a number of very clever machines that could in some ways imitate human behavior. So, here he's got a statue which has a fire in it, you light this fire, I'm trying to point to the fire here. You light this fire, and it boils water reservoir of water, that's inside the machine causing steam to expand, and then causing that figure at the right to lower its pitcher of water and douse the statue. So this is a statue that puts itself out. It's a very clever device. The statue itself is no longer in existence, but we have evidence that Hero created or designed such a statue. This is already a kind of intriguing thing. It suggests that people are thinking about machines that can provide the illusion of having purpose or human behavior. Over the years, the technology involved in creating these machines got more and more sophisticated. By the 1400s you were seeing clocks in Europe. This was really the the area of technology where sort of fine-tuned machinery made its greatest advances. So, over time as people started to make clocks, they started to embellish them with lifelike figures that could do things like strike the hour. The figure at the upper left here is a clock striking person, they were called Jacks of the Clock in England. And and this was made in about the 1400s, and from what I know it's still operational. Over time over the next few centuries, the art and technology became even more sophisticated. And unlike the kind of public civic architecture of clocks, advanced machines automata began to be associated with objects of luxury art made for aristocrats, sometimes often in fact embedded with clocks. But at the same time, their appeal was just that they were spectacularly well designed and extraordinarily life like. Maybe the high point of the history of automata, of life like machines, is given by the the picture at the bottom here. The great genius of automaton building was a man by the name of Jacques de Vaucanson who in France in the 18th century. And by this time automata were starting to be not just for the aristocracy, but they were starting to be objects of display, almost like special effects. You almost get the feeling that these displays of automata were kind of like a summer blockbuster movies, or something like that. The automata that are shown in this picture, again, none of them are still in existence, but the automata that are shown in this picture are three of de Vaucanson's masterpieces. The one at the left is a flute player, and that's an actual flute in its hands. So the machine was actually able to blow air across the top of a flute and get notes out of it, it's kind of extraordinary work of engineering. Vaucanson's most famous piece was that little duck in the center, people used to come owe and art this duck. You would place metal pellets in front of it and the duck would lean over and eat the pellets, and you would see its feathers ruffle and then it would excrete. And this was just seen as the most amazing piece of lifelike art ever. The reason I go through these things is because these are instances of machines in people's experience becoming more and more lifelike. I should say human-like except in the case of the duck, maybe it's more duck-like. But regardless, these are machines that are becoming more and more imitations of life. That causes people at the same time to start wondering if human beings can be treated as machines. Just as we can say that a machine is sort of like a person, maybe we can also say that a person is a little bit like a machine. And as people start to think about this, it's actually a rather provocative and worrisome thought, but you start seeing, for example, in the 1500s, the work of a surgeon, a French surgeon named Ambroise Paré. He was a battlefield surgeon, he was interested in making prosthetics. And so, I don't actually know if the diagrams in this page of his book represent things that were in fact actually built. I'm not sure about that. But again the fact that he was modeling and conceiving these things, is quite interesting. To make a prosthetic hand, you are already imagining that the workings of the human hand can themselves be modeled by a machine. In other words, now we're not talking about machines that are lifelike, we're talking about humans starting to be viewed as machine like. Rene Descartes, the 17th century philosopher, began actually, he pushed this idea a little further, and he was aware that it was a fairly risky and to some heretical idea. But in some ways he was intrigued by the idea of modeling human behavior in machine terms. This is a very famous diagram where he displays his idea of how reflexes work. So, you'll notice that the that in broad outlines the diagram doesn't look too unlike that Hero of Alexandria statue. The idea is that if a person puts their hand near a fire, then there is some kind of perhaps pneumatic pressure that runs up the lines of nerves to their brain, and causes them like a machine, to pull their hand away from the fire. That's, again, in some ways a very daring and risky notion of thinking of human behavior as being machine-like. Descartes famously stopped short, however, of thinking of the mind as a machine. But over time even that line started to be broached, and the notion that people could be thought of as machines began to be voiced more and more openly. There's very famous but to me rather confusing book from 1747 by Julien Offray de la Mettrie called Machine Man, and I'll read this quote out to you. Again, this is a typical quote from the book, and it's typical in that it's vague, and reading the entire book I don't think is of great interest unless you're interested in the history of these ideas. But it's a difficult book and it seems like de la Mettrie is concerned throughout with pushing the boundaries of how far you can call people machines. Simply admit that organized matter is endowed with a motive principle which alone distinguishes it from unorganized matter, and that an animal's everything is dictated by the diversity of this organization, as I have sufficiently proved. That is enough to solve the riddle of the substances and of man. We can see that there is only one substance in the universe, and that man is the most perfect one. He is to the ape and the cleverest animals what Huygens planetary clock is to one of Julian Laroy's watches. Again, there's the use of the clock as a metaphor for an intricate mechanism. So far, we've talked about machines being seen as being lifelike and humans being seen as machine-like. And after the 1700s, the lines between the two begin to blur even more in literature and in philosophy of science. A wonderful example from the early 1800s is a short story by the German writer ETA Hoffmann, story is called the Sandman. And it introduces this kind of eerie machine named Olympia, a machine woman who is so lifelike as to be strange and and disturbing. I have a quote from the story at the bottom of the slide there. It's a fantastic story to read, and all the more fantastic because it begins to point at the at the anxiety that people feel around machines that can behave like people. The anxiety was somewhat dispelled later in the century when the story was turned into an opera by Jacques Offenbach. And what you're seeing on the screen is a picture of a modern production of Tales of Hoffmann,where Olympia the doll sings the song. Offenbach played this for laughs, but the Hoffmans story is anything but humorous, it's a it's a very serious and scary story. It's an example of the kind of conversation that began to be felt more and more. The anxiety around machines as people can't be denied, and it's going to be something that in part we're going to deal with as this course goes on. It's reflected in the 20th century in too many to count stories involving machines that act like people, that rebel against people, that fool others into thinking that they're people. The first use of the word robot was in a play by the Czech playwright Karel Capek called RUR standing for Rossum's universal robots, that was written in 1920. But all through the century there have been numerous examples of machines that make people nervous. Robots, androids, computers later on in the century as computers arrived, computers run amok, and so forth. And we haven't evaded this fear to this day, this is just one example, robots can be scary machines can be scary. I just point to one little thread of this sort of cultural progression in which things like dolls and ventriloquist dummies can be scary. So, there are things like well the picture up at the left there, top left is from a movie called Dead Silence, with a pretty scary dummy. And recently there have been a series of Annabelle movies with the scary doll. My favorite of these is the one that's at the bottom a British movie from 1945 called Dead of Night. You should really find this movie and watch it. I'll make that an informal assignment for the course, find Dead of Night and watch it, it's great. It's really scary, and even if you don't like horror movies generally, it's not gory or disgusting or anything like that, it's just scary. But it features, it's a bunch of stories, and the last story in it features a ventriloquist and his dummy, and let's just say that the relationship between the two of them is not entirely amicable. Dolls and machines and lifelike devices can be anxiety-provoking to us. And that's a theme that's that in fact is actually going to run through a lot of the discussion that we're talking about. People have an emotional investment in whether humans can be modeled by machines, and how accurate machines can be as models for human behavior. We are going to talk much more about this as time goes on, but just to wrap up this one discussion. We've mentioned the idea that machines can imitate humans in an effect that's the bedrock foundation of the field of computer science known as artificial intelligence. Artificial intelligence is the study of trying to get computers or machines in general to imitate what we would think of as intelligent behavior in humans. Similarly, bedrock foundation of cognitive science, which we will also be discussing here, is the idea that humans can be modeled adequately by machine programs. We can learn about how humans behave by creating programs or devices that illuminate or imitate their behavior. And finally, they're still frontiers that are yet to be explored involving the integration, the weaving together, of humans and machines. We will be talking about this toward the end of the course, but I mentioned it now just because it's an area of some interest to me. The picture that you're seeing is an Australian performance artist named Stella Arc. And the in the picture he's shown wearing a robotic arm that is in fact attached to his arm through electrodes. He can control the arm by or he, I don't think he does this particular act anymore, but when he was wearing the arm he could control it by muscles in his stomach. And he became a facile enough with the arm so that he could do things like write his name, for example. That's an example of weaving together human and machine intelligence that is not quite identical to either of the two or more straightforward things that we talked about so far. But these are going to be the directions that we are talking about, provocative directions as we move on in talking about cognitive science and artificial intelligence.
