Re-creating Ourselves, Part 2

God in the Machine: What Robots Teach Us About Humanity and God
by Anne Foerst, Ph.D.

(Page 3 of 3)

How Humans Bond with Machines

Imagine yourself sitting in a lab in front of a computer and talking to it. You chat with it, and it reacts and asks good questions. Your conversations become deeper and deeper and you finally find yourself discussing with the computer private and intimate problems. You write down how your problems with your partner have started to threaten your career and how daily problem-solving is demanding too much energy from you. You tell the computer that your partner is a control freak, that you are secretly drinking, making yourself more dependent and helpless without any way out.

The computer seems to express sympathy, asks specific questions, and empathizes with you, and you have the feeling the computer understands you better than anyone else. But the underlying program depends on an outdated therapy model in which the therapist just mirrors the comments of the client and turns them back to the client in the form of a question; e.g., "I hate my father!" "Why do you hate your father?" This very simple behavior gives the computer user the feeling that the objective and rational machine sympathizes with her, so she cannot be all wrong.

Does this scenario sound unlikely to you? Well, these very scenes have occurred since the 1960s and are happening today all around the Internet-accessing world. The first program of this kind, ELIZA, was programmed by Joseph Weizenbaum, then professor of computer science at MIT, who was a major inspiration for my doctoral work. ELIZA was based on the mirror method and had, in addition, a couple of standard questions the program could ask if it didn't figure out the keywords of the client's comment. When I met Joe in Germany, I was studying theology and just finishing my degree in computer science. I heard him giving a lecture on the dangers of computers. He told the story of implementing ELIZA in the '60s and then how he observed some of his graduate students, who should have known better, using ELIZA as their own personal therapist; later, versions of ELIZA were and are frequently used in some initial psychological analyses.

We can argue that this sounds very naïve and that the computer-savvy users of today would rarely fall into the same trap. Well, this optimism is misplaced.

In my travels, I met Cliff Nass, a sociologist at Stanford who analyzes how people interact with machines. Have you ever screamed at your computer or kicked it? Have you ever complimented it? Cliff wanted to find out if there is any evidence that we treat computers differently than typewriters, if there are any emotions involved when we interact with our desktop.

For one of his earliest experiments, he asked several people to test a computer-learning program that was supposed to be introduced into elementary schools. The program was very bad. Some of the testers were computer specialists and some were laypeople. After they had tested the program for a while, the computer on which they worked asked them to evaluate its performance. For the most part, people responded positively. Afterward, these same testers were led into another room with other computer terminals and were asked to evaluate the learning program again. Here, on these different computers, their answers were less positive about the quality of the tested software but they still sounded somewhat satisfied. Finally, a human asked the testers for their opinion on the software and the testers were very negative about it. Such a program should never be used in school, they said.

Interestingly enough, the testers had not voiced these criticisms to either the computers on which they had tested the program or the computers on which they had done the second evaluation. These same people, when asked if they would ever be polite to a computer or think they could hurt its feelings, vehemently rejected such a notion.

This experiment suggests that somehow we seem to apply our rules of politeness to nonhuman entities such as computers. Obviously, the participants in the experiment did not want to hurt the computers' feelings. They even assumed a level of kinship between different computers and, therefore, applied similar rules of politeness to the computer on which they did the second evaluation. They didn't tell these machines their true, very critical opinions, either out of the desire to not hurt the feelings of the second computer by criticizing one of its "fellow" computers or because they assumed some contact between the two so that the second would tell the first what had been said. It seems that somewhere during our interactions with a computer we start to assume that a computer is as sensitive as a human being. Therefore, we behave politely and don't want to criticize it openly.

Let's look at another experiment that seems to imply that people bond with their computers. In this case, Cliff placed people and computers inside one room. Half of the computers had green monitors while the other half had blue monitors. Half of the people wore green arm badges; the other half wore blue ones. Together, they all played interactive games, and it turned out that the people with blue arm badges were much more successful using computers with blue screens to reach their goal than using "green" machines. The same, of course, was valid for the other side. So, slowly, the people with green arm badges bonded with the green-monitored machines and the blue-badge people with the blue-monitored machines.

And now comes the surprising result. After approximately half an hour, the people wearing the blue arm badges expressed more solidarity with the computers with the blue screens than with the humans with the green arm badges; the same was true for the humans with the green arm badges. It seems that through the interactive games and the experienced benefit of interacting with the machines with one's color code, the color code took over as a definition for "my" group. The entities with the other color code, no matter if humans or machines, tended to be rejected. Remember the fictive fight between two groups in front of the vase-face gestalt image, with one insisting that the faces are an illusion and the other insisting the opposite, that the vase is an illusion. What happened in this second Nass experiment is that same thing, only this time the groups had nonhuman members and the group was subtly forged through the interactive games.

This seems to imply that humans bond with the entities of their own group, whether they are human or not. As we will see later in the book, humans do not have a sense of kinship with other humans "built in." It is not part of our biological makeup to automatically treat all humans better than all other beings. That means we bond easily with nonhuman entities and, therefore, we bond easily with our computers.

Human beings are social mammals. Most of us seek places where we can meet other people. We go to bars and sports events and demonstrate with other like-minded people. We naturally want to be with other people. And, as Cliff's second experiment shows, we seem to be able to accept anyone or anything into our group with whom we can sufficiently interact. As soon as such a stranger is accepted into a group, he or she is seen as an equal part of the group. The group defines itself by the entities that both belong and do not belong to it. As we all know and perhaps do ourselves, people treat their cars and stereos as people as well. In a way, it saves a lot of time and energy to do so. After all, humans are educated from birth on how to interact with their fellow human beings. It is necessary for a baby to be able to do so, as her survival depends on it. Throughout our lives, we learn patterns of behavior-such as being polite and not openly criticizing someone. It is very easy to apply these ingrained rules to every entity we interact with. It is very hard not to do so, as it demands a conscious effort of us.

This behavior-treating nonhuman objects as if they deserve some form of politeness or regard and are somewhat like us-is called anthropomorphism, the human capability to morph/change everything into a human (anthropos) and treat it accordingly. Usually, the term has a slightly negative connotation. Theologians especially often criticize human terms used to describe God, such as shepherd or father, or the classical image of God as an old, usually Caucasian man with a long white beard. Cliff and his colleague Byron Reeves, however, suggest that anthropomorphization is actually the initial and natural response to anything we interact with; it takes a conscious effort not to anthropomorphize. As social mammals, we are best when we interact and any use of these trained and built-in behaviors is easy; anything else is hard.

Cliff's experiments reveal something about how we relate to one another. We innately work toward community. We work toward social interactions. We work toward groups. We work toward bonding and solidarity with the beings closest to us. And computers are great thinking tools we can use to explore our mechanisms for bonding and social interaction.

The First Humanoids: What Happens When We
Do Not Bond with the Creatures We Create?

Humanoid robots are fascinating machines and there are many science-fiction movies and popular stories that deal with them and how they shape the way we live and think about ourselves. Our culture is filled with stories about humans who lose power over creatures they have created. In the context of computers and robots, it is not just HAL from 2001: A Space Odyssey that raises these fears in us. There is another, even more powerful story that had a major impact on our reactions to computers and particularly to robots.

Mary Shelley's Frankenstein is about a monster that became cruel and turned against his maker, his family, and society. Popular culture today sees hubris as the main motif of the story. The concept of hubris comes from Greek mythology and addresses the capability of humans to overestimate themselves. It is not just arrogance but more the feeling that humans can do everything. In antiquity, the prototypes for hubris were Prometheus and Ikarus. In Frankenstein, the prototype of a hubristic human is the scientist Frankenstein who builds a creature that ultimately will go against its creator and his/her family and utterly destroys them. This motif touches a deep fear in us: that our own creatures will turn against us. This novel has influenced our thinking about robots so much that authors such as science-fiction writer Isaac Asimov called the fear of humanlike machines the "Frankenstein Complex."

But I see another story. A young and overeager researcher named Frankenstein builds a creature out of human parts but then gets scared when the creature comes alive. He runs away and abandons it. Thus, the creature is never given a name, never treated as a beloved offspring. It tries to find community but is constantly rejected and feared. It is completely ostracized. But when a humanlike being has no chance to actually become part of a community, it seems reasonable that out of grief and revenge this being might turn against the community that excludes it. The monster's motive to kill did not evolve out of a need to exert its power over others; it stemmed from the feeling that it was not accepted by the community. How could the monster ever develop any form of benevolence toward humans when all it experienced from them was hatred and rejection? It never experienced positive feelings of warmth or kinship. It became brutal because it didn't know what else to do. No one let it participate in any form of human interaction. There was no bonding.

Most Frankenstein movies cover only the aspect of hubris and destruction, with the notable exception of the movie Mary Shelley's Frankenstein, where the creature at some point challenges Frankenstein. After helping a family anonymously, only to be rejected when it is discovered, it asks Frankenstein, "I can read, I can learn, I can think. Do I have a soul or have you forgotten to build it in?" Here, the creature clearly desires much more than just being self-aware and the capability to understand. It wants to be humanlike, in the sense that it wants to be accepted and loved. It wants to interact. It is so much like us that it wants to be connected to a community. In this story line, Frankenstein is about responsibility toward the creatures we create. We have to care for them and treat them well; otherwise, there might be unforeseeable consequences.

The Frankenstein story is, of course, not the only one that talks about the possible outcomes when we create creatures in our likeness. Much more interesting theologically and, thus, much more relevant to this book are the stories about the golem.

Golems

The wife of Rabbi Löw could not understand why her husband had forbidden the use of the Golem for private purposes. And when, just before Passover, she was short of help she allowed herself to give the Golem orders to fill two large water kegs which stood in the kitchen which was all prepared for the holiday. She thought also that a service in preparation for the Passover feasts did not come under the head of secular purposes.

But she had a very unpleasant experience.

The Golem took the pails and ran swiftly to the brook.

Several hours later the courtyard of the house of the Rabbi was flooded with water, and people were crying: "Water! Water!" The secret source from which this water was flowing was sought.

But it was not found until the Golem was seen patiently obeying his orders by continuing to pour water into the kegs which had been filled a long time before. This explained the flood and there was much laughter over the Golem's mistake. [ . . . ]

Since that time the people took care not to give the Golem any profane work to do. To this very day in Prague people say to an unskilled artisan: "You are as competent for this work as was Joseph Golem as water carrier!"

Chayim Bloch, The Golem: Legends of the Ghetto of Prague, trans. Harry Schneiderman (Blauvelt, N.Y.: Rudolf Steiner Publications, c. 1972)

The Jewish golem stories go back to the thirteenth and sixteenth centuries. They can be traced back to medieval Germany and Hungary, specifically to the Jewish mysticism, called kabbalah.

The term golem comes from the beautiful Psalm 139 quoted at the beginning of this chapter, which is one of my favorites. It talks about the beauty of creation and how God knows every one of us and knows everything in us. Even if the psalmist sometimes finds this omnipresence of God somewhat disconcerting, it also gives her deep trust, as she is sure that God will never leave her.

The verb galam appears only twice in the Hebrew scriptures. In 2 Kings 2:8 it is used to describe the wrapping of a mantle. But probably the oldest source for this term is in Psalm 139:16: "you created me as a golem in my mother's womb." Here galam is usually translated as "shapeless thing" or "embryo." The psalm celebrates creation and the special love and care of God toward humans. God created the psalmist, "intricately woven in the depths of the earth," and in God's "book were written all the days" that were formed for the psalmist. The word galam very likely comes from an Arabic root and means "tangle" or "cluster." The medieval kabbalists used this term as the name for the humanoids they constructed.

The most famous golem story is the story of Jehuda Löw, the Maharal of Prague, and his golem, Joseph. The Maharal is a historical figure who lived in the sixteenth century. He was a widely acknowledged theologian and also a political figure. He was a very influential teacher and a very wise negotiator with the Christians and the state representatives to create a decent life for the Jews in the ghetto.

At the time of Rabbi Löw (as, unfortunately, in most of medieval and even modern times), Jews were often attacked by Christians, and the people in the ghetto of Prague were often harassed. So, to add a layer of protection to the ghetto, Rabbi Löw is supposed to have built a golem and put a paper with God's name in its mouth. The golem then became animated and was able to help the Jews in Prague. Joseph supported the Jews with his strength in their daily labor and helped them against attacks from outside. One story describes how Christians would hide dead babies in the ghetto at night and then come back during the day with armed forces, and use these little bodies as proof that Jews would kill babies in their ceremonies. Then, Christians would have a reason to attack the ghetto and kill Jews. The golem is known to have found the babies several times and hidden their bodies so that the accusations became worthless.

According to most stories, golems are built from clay, constructed through words and numbers. Kabbalist theory reveals a deep faith that the world was created by God in an orderly and numeric fashion; the better people understand the logic behind the world, the more they can share God's mind and participate in God's creativity. Thus, they are motivated to construct increasingly complex things to understand God better. But they cannot build anything animated without help; golems come to life only if they have a paper in their mouth with the holy name of God written on it, or with God's name engraved on their forehead. The ultimate power of life is God's and God's alone; God has to be involved to animate an artificial being. So, even if the letters and numbers in Hebrew are orderly and thus participate in the order of God's creation, they are not sufficient on their own to create life. Quite the contrary: The tangle of flesh, genes, slime, and chemistry in the case of the human animal, or the clay in case of the Golem, need the spirit and power of God to become alive.

One could now ask if the Golem is estranged, a sinner. According to most stories, the Golem has no language and thus cannot participate in the categorizing, describing, and reducing of facts. He also has no sense of right or wrong-otherwise why would he flood the Rabbi's house? But there is a small potential that he also needs God's forgiveness. Rabbi Löw himself was never sure if the golem was a child of God or a mere machine. Since the Maharal cannot be absolutely sure, Jehuda Löw addresses this doubt by forcing the Golem to keep the Sabbath. Every Friday, the Rabbi would remove the animating paper with God's name on it from the Golem's mouth so that it went back into its unanimated state, thus keeping the Sabbath.

One week, however, the Rabbi forgot to remove the paper slip and the Golem, without his master, went berserk. Rabbi Löw saved his fellows of the ghetto by fighting the Golem and, after considerable violence, he was finally able to remove the life-giving paper from the Golem's mouth. In some versions of the legend, the dying Golem falls on the Rabbi and smashes him. These endings refer to the motif of hubris, as often presented in Greek tragedy and also in the Frankenstein story, where the constructors of gadgets and creatures that overcome human limitations are killed in the end.

Golems as Prayers

What is prayer? Is it to ask God or another deity for something you want, such as money or health? To have a conversation with yourself? Meditation? Do we pray when we ask the deity for the strength to accept the situation we are in, to endure what we can't change in our lives?

In the Jewish context in which the golem stories are told, prayers are usually spoken to celebrate God and God's glory in us. Prayers are communal; they strengthen the bonds of people in their community, with each other, and with God. People speak prayers to express their anger, fear, and frustration with current situations, as well as to speak about their joy in life and their happiness to be God's creation.

With the construction of golems, people felt they learned more about God's creation of humans and their special capabilities. The golem builders felt that by building golems they were participating in God's creativity. They argued, God has created us in God's image so that we participate in God's creative capabilities. Whenever we are creative, we actively participate in God's creative powers and celebrate God. In this sense, every act of creativity is a prayer. And the more complex things we build, the more we praise God. Humans are-at least as far as we know-the most complex beings on earth. Therefore, if we rebuild ourselves in golems, we celebrate God's "highest" creative act, the creation of humans, thus praising God the most.

Even if one ending of the story of the Maharal in Prague refers to an element of possible danger, the majority of the golem stories are not about the motif of hubris as the Greek stories or the Frankenstein myths are. This is supported by a vast amount of rabbinical literature that discusses golems. The majority of these stories do not understand the construction of golems as a step beyond the boundaries God has set for us or as hubristic acts, but they understand golem-building as prayer.

Golems can be helpful servants, but their creation has a spiritual purpose beyond building useful machines. It can be an act of worship. It's not trying to dehumanize the human experience or deconstruct the mystery of what it means to be human. It is to praise God. This, of course, links these golem stories with the modern scientific construction of humanoid robots.

While I was at MIT, the group that was constructing Cog and Kismet had just made the jump from insectlike, six-legged walking robots to humanoids; while they all enjoyed trying to do the impossible, most of them were also well aware of the challenging nature of this task. Everyone was enjoying him or herself trying to turn science fiction and mythology into reality. But many also felt a nudging doubt. After all, they had had experiences with the rebuilding of insects. And even if the six-legged creatures they had built were fantastic robots, they came not even remotely close in their capabilities to real insects. Now, they were building a humanoid and realized that they might be out of their league. Each week for the first year of our work, we invited a specialist to tell us something about how the human system works. We would read their articles beforehand and pepper them with questions after, and one question we asked every single one of them was: How will this knowledge help us to build our humanoid? The more we learned, the more our respect grew for the incredible complexity of the human system. It is one thing to read the Bible to learn that we are the "top" of creation, but it is quite another to learn the facts about this complexity. We became admirers of ourselves, but that didn't make us arrogant. Quite the contrary, we now had a healthy respect for the task we had set ourselves, to build a system that has, however remotely, some similarity with us. Building Cog and Kismet made us modest in our admiration for God's creation. Probably no one except me in the team would formulate this feeling with the same religious words, but the sentiment was exactly that. Modesty, admiration, and a humble attempt to do our best is all that we can do to make it work.

This experience has taught me the truth of the view of the kabbalists to see the construction of humanoids as a worship of God.

It also says something about God's creativity in us that we don't stop our projects when we discover how difficult they really are. It takes a healthy self-confidence to attempt the impossible, and even if you fail, you have tried. Cog and Kismet, though both now in museums, have contributed hugely to the field of robotics, as they were the first emotional and social robots. Even if we have the capability to bond with most creatures with whom we interact, we bond most easily with those creatures that seem to have emotions that we understand. Anthropomorphizing happens with dogs and cats and other mammals because their basic similarity to us motivates us to create stories that ascribe to them the same emotions we have, and to bond with them strongly. Cog and particularly Kismet evoke similar feelings in us and thus have moved us closer to the eventual fulfillment of the old human dream of rebuilding ourselves.

One of the oldest golem stories tells about Jeremiah and his son who built a golem with the words JHWH elohim emet (God the Lord is truth) on its forehead. As soon as it came to life, the golem erased the letter 0 (the aleph, the first letter of the Hebrew alphabet) from the word truth so that now his forehead said JHWH elohim mot (God the Lord is dead). He then explained to his terrified builders that we adore God because God has created us, the most complex beings there are. If we are now able to re-create ourselves, people will adore the constructors of golems and not God anymore. But a god who is not adored and prayed to is dead.

One has to be honest and admit that this aspect of golem construction is always present. Even if we become modest and admiring of creation, there is still often the sense of demystification. If we learn a scientific explanation for a unique human capability, for example empathy (see chapter four), we are tempted to lose respect for this capability, as it seems to be just a reactive mechanism. The construction of humanoids certainly contains an element of hubris. The pursuit of an impossible task produces pride in one's own work and in the triumph of overcoming limitations, to succeed when no one in the scientific community thought it possible. There is nothing wrong with that. Quite the contrary, those elements are normal in any creative endeavor. But there is always the danger of losing respect for the human system because when you build robots and they do what you actually want them to do, you sometimes can't help but think that ultimately every part of the human system will be understood and, then, can be rebuilt.

The construction of humanoid robots is motivated by the wish to understand ourselves and to build partners with whom we can talk and interact in a meaningful way. If they challenge our self-understanding or seem threatening to us, this is not the creatures' fault but the outspoken hubristic aspects of this project that are sometimes part of humanoid projects.

Theology and Artificial Intelligence meet when we attempt to understand ourselves, who we are, and what our role in this world is. Engineers and computer scientists involved in Artificial Intelligence, as well as theologians who (re)construct stories of meaning, are creative. Both can be described with the metaphors of Homo faber and Homo narrans. The first attempt to rebuild themselves; the others tell stories. That is, both contribute equally to a fuller understanding of ourselves. They do professionally what many humans do in their spare time: attempt to answer the big question of what it means to be human.

Golem Builders Today

In true rabbinical fashion, there is more than one version of the story of Rabbi Löw and the golem Joseph. We know already about one ending, where the lifeless golem, without the animating paper in its mouth, falls on the rabbi and kills him. We have seen that this ending shares with the Christian tradition the element of hubris present in many projects of humanoid construction. I would like to focus now on the tradition that tells the story of the survival of the rabbi. In this version of the legend, the golem is put to rest in the attic of the synagogue in Prague. Rabbi Löw then creates a kabbalist rhyme that will revive the golem at the end of all days. Many Jewish children from this tradition were taught these words.

This version of the kabbalist golem legend is still strongly ingrained in the consciousness of many Jews from the Eastern European tradition. When Jewish boys descended from Rabbi Löw were bar mitzvahed, they were usually told the formula that will revive the golem at the end of all times.

It seems that many of the early AI researchers are or claim to be descendants of Rabbi Löw, but it took a coincidence to find that out. MIT is the cradle of AI; here, the field of AI was born and here the first steps toward artificial intelligence were taken and the first successful projects developed. In the late 1960s, when some students sat together on a break, someone mentioned that the first big computer in Rehovot, Israel, had been called Golem. This led to a discussion and it turned out that at least two students in the community had been told the rhyme that would awaken the golem. These two were Gerry Sussman, today professor at the MIT AI Lab, and Joel Moses, the former provost and today institute professor at MIT. When they compared the formulas they had both been told, their formulas were exactly the same-despite hundreds of years of oral tradition.Gerry Sussman later dedicated his doctoral thesis to Rabbi Löw because the rabbi was the first one to recognize that the statement "God created humans in God's image" is recursive. Recursive functions are self-referential; that is, one cannot derive all values individually but needs the previously calculated values in order to get new values. This dedication captures various aspects of the AI enterprise. For one, God has created us in God's image and we use the same process in humanoid construction as we create them in our image. Modesty and awe come out of humanoid construction, as we can never be as successful as God. We are a derivation of God and our creatures will be the next derivation, our images. To interpret the imago Dei as recursive also refers to the aspect of prayer, as we can create only because we have been created in the first place and celebrate our creator who has so "wonderfully made us" (Psalm 139).

It also points out the necessity to re-create ourselves. We are images of God and we have the drive to create, to repeat God's acts of creation. The very desire of God to create humans as partners is inside us. When we look at all the attempts to "speak" with animals, especially dolphins and chimps, and the desperate search for extraterrestrial intelligence, it becomes clear that, for some reason, humans want to interact with beings of a different kind. We want to have a species, an "other," with whom we can interact. We know that many other animals are intelligent but we cannot communicate with them. But there is hope that we can communicate with the beings created in our image. They have the potential to be partners.

Recursive functions have yet another attribute, as one cannot derive a value in a recursive function without having calculated the previous value. Does this mean God needs us in order to create humanoids? Has God perhaps created us for this very purpose? Why else this strong and so deeply ingrained desire to re-create ourselves? These delightful speculations add another aspect to the element of prayer within the golem tradition. When we attempt to re-create ourselves, we do God's bidding. We help God. We are created co-creators, a term coined by the theologian Phil Hefner, who has influenced my thinking in religion and science enormously.

One might further speculate that the wish to revive the golem at some point in time is part of the motivation for the whole AI enterprise; this seems especially to be true, as several other famous AI researchers link themselves to this tradition.

In light of the state of sin and ambiguity in which we live, golem creation has a forward-looking perspective. As we have seen, it is doubtful that the golem was living in ambiguity. But here another spiritual aspect of the AI enterprise is revealed. Wouldn't it be wonderful if we were to create an image of ourselves that is not a sinner, and can actually handle the ambiguities of life? Can we actually build creatures that are better than us? Not smarter but morally better?

Those of us who have read the robot stories of Isaac Asimov know that he indeed presented his robots as better people. They could do no wrong, as they had to obey the three laws of robotics.6 Asimov presented one of the most powerful examples of our creations that is morally perfect. But in his later books, he realized that these laws of robotics fall short in a universal context. He, then, adds to the three a "Zeroth" law that says "A robot may not injure humanity, or, through inaction, allow humanity to come to harm." In this moment the robotic laws become as ambiguous as human laws, as the term humanity is ambiguous. Who belongs to humanity and who doesn't? If one part of humanity decides to kill all the others, can that be accepted?

Many people do not like speculations like these. Many even reject a connection between the golem stories and modern AI. Many people, especially scientists, do not want to acknowledge the existence of such emotional and religious elements in the motivation of researchers, nor do they feel these elements are desirable. However, when we look at scientific enterprise today, we realize that there is much room for researchers to bring in their own quest. Many robot builders bring to the table their desire to see what humans can accomplish with the help of the technology they have available today. Many feel that robot-building can be spiritual, as it taps into God's creative powers in us. Many people might be motivated by the inherent desire to overcome human flaws and limitations by constructing robots that are better or free from sin as estrangement.

Only if we see the enterprise of developing artificial intelligence as purely scientific and ignore all the mythical and emotional elements will we be in danger of falling into the trap of hubris. Therefore, we have to first overcome the danger of seeing AI as a purely rational endeavor before we can actually look at it as a source of wisdom about who we are.

1. The various layers of the Hebrew scriptures use several names for God. The oldest one is the acronym JHWH, often falsely pronounced "Jehovah." Around 800 bc, the Jewish community, in fear of trivializing the name of their God, stopped saying the name aloud; whenever the tetragram appeared in the biblical texts, they would say "adonai" (My Lord), or "elohim" (God). Out of respect for the Jewish taboo, many translations of the Hebrew scriptures translate JHWH with "my Lord." For easier reading, I will translate the acronym JHWH simply with "God."

2. One concrete example for the application of this principle would be suicide. Even if I judge that it is the right decision for me, I could never claim it has any universal value! If every human being would commit suicide, that would be the end of our species and, therefore, unreasonable and unacceptable.

3. These "deadly sins" are not part of the Roman Catholic catechism; instead, they have been developed by the people themselves and, then, have become part of Roman Catholic folklore.

4. Letters in Hebrew are also numbers-there are no special numeric signs. This gave the kabbalists the opportunity to create word and number riddles far more complex than our Western languages allow for.

5. Among those people who have been told the formula are John von Neumann and Norbert Wiener, both foundational thinkers in the field of AI.

6. Third Law: A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Second Law: A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

First Law: A robot may not injure a human being, or, through inaction, allow a human being to come to harm.

© 2004 Anne Foerst, Ph.D. All rights reserved. No Part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any other information storage and retrieval system, without the written permission of the publisher.

About the Author

Dr. Anne Foerst is a former research scientist at the Artificial Intelligence Laboratory at MIT, where she also founded and directed the God and Computers Project. The only robotics theologian in the country, her work has captured much media attention, including coverage in The New York Times, The Boston Globe, and Science. She is currently a visiting professor of theology and computer science at St. Bonaventure University and lives in Olean, New York.