Direct Contact with Extraterrestrials via Computer Emulation
by
William Sims Bainbridge
National Science Foundation*
Forthcoming in Culture in the Cosmos
Edited by Douglas Vakoch and Albert Harrison
*The views expressed in this essay do not necessarily represent the views of the National Science Foundation or the United States.
ABSTRACT
This essay suggests that cognitive science and information technology are progressing in a direction that will make it possible to transmit realistic avatars of individual human beings to extraterrestrial civilizations. The first step is personality capture, the process of gathering sufficient information about a human being’s memories, thoughts, and feelings to allow emulation of the person in a computer, information system, or robot. Personality capture must employ culturogenic measures, because personalities arise in a particular cultural context. Once recorded, personalities can be delivered to the stars in Starbase interstellar database probes or via radio given the willingness of a recipient civilization. Embassies from multiple civilizations can be situated in a virtual reality Cosmopolis, where computer-generated avatars emulate the original biological beings in social interaction, trade, and cultural innovation. The chapter ends with ideas about practical steps we might undertake to fund, develop and apply the needed technology.
Direct Contact with Extraterrestrials via Computer Emulation
Radio transmission of human personalities to other stars is a logical corollary of personality capture, the process of gathering sufficient information about a human being’s memories, thoughts, and feelings to allow emulation of the person in a computer, information system, or robot. Steady progress in artificial intelligence, cognitive science, and personality capture raises the very real possibility of transmitting functional avatars of human beings to corresponding extraterrestrial civilizations, and vice versa. These avatars would be high-fidelity autonomous agents capable of perceiving, thinking, and acting in manners very similar to the humans and ETs on which they are based.
Historical Background
When I first seriously proposed personality capture in 1993, I did so in the form of a parable about a fictitious engineer who had been working on a NASA SETI project to detect radio signals from extraterrestrial civilizations. In the story, he undergoes an experience of almost religious revelation, standing on the actual surface of the great Arecibo radio telescope, realizing that he personally could travel to the stars if he scanned his personality into a computer and had it sent, perhaps centuries later, on an interstellar probe. After eight years of research, I presented a better grounded version of the idea at a NASA conference:
We have the technology, already today, to begin archiving human personalities at low fidelity within what I call Starbase, a database destined eventually to be transported to the stars. To gain entry to Starbase, a person must contribute significantly in some way to the creation of interstellar civilization. One way is to help develop technologies for archiving and reanimating human personalities at ever higher fidelity. Another is to work toward the establishment of small human colonies, first on the Moon and Mars, where Starbase can be headquartered and where serious work on reanimation can begin.
When the time comes for the first interstellar expeditions, they will be carried out not by biologically-based humans in their first brief lifetimes, but by eternal Starbase modules incorporating the archived but active personalities of the crew and colonists. At the destination, the crew will not waste its time terraforming planets, but will adapt the colonist into whatever form (biological, robot, cyborg) can thrive in the alien environment. Subsequent waves of colonists can be sent as radioed datafiles, in a technically feasible version of the old science fiction dream of teleportation. (Bainbridge, 2002b: 61-62)
I am sure that many visionaries and science fiction writers have had similar ideas. Back in 1966, Roger MacGowan and Frederick Ordway argued that interstellar explorers will be robots rather than biological organisms, for example. Recently, Steven J. Dick (2003) has argued that the dominant forms of intelligence in the universe may be “post-biological,” raising the question of whether our own species is ready to transition away from its traditional biological form.
The crucial point is that the convergence of cognitive science, uniting cognitive psychology with artificial intelligence and cognitive neuroscience, has prepared a basis for rapid and very practical progress in personality capture and emulation. In partnership with Mihail Roco, the charismatic leader of the National Nanotechnology Initiative, I have been able to organize a series of scientific conferences exploring the convergence of nanotechnology, biotechnology, information technology, and new technologies based in cognitive science. At the first of these Converging Technologies conferences, sponsored by the National Science Foundation, computer graphics pioneer Warren Robinett projected the consequences of understanding how the human brain actually works. He noted:
If a mind is data that runs on a processor (and its sensors and actuators), then that data – that mind – can travel at the speed of light as bits in a communication path. Thus, Mars is less than an hour away at light speed. (We need a rocket to get the first receiver there.) You could go there, have experiences (in a body you reserved), and then bring the experience-data back with you on return. (Robinett, 2003: 169)
Robinette and I both assumed it would be necessary to transport a receiver and many support systems to the destination, but that would not be the case if an extraterrestrial civilization were willing to serve as host for interstellar informatic travelers – infonauts. With that possibility in mind, we can consider what actually would be involved in capture, transmission, and emulation of human personalities.
Personality Capture Methodologies
To introduce personality capture efficiently, I will explain my own personal approach to it. My family has always placed great emphasis on preserving personal historical records, and my entire life has been spent surrounded by ancestor’s publications dating back to 1856, photos from the same decade, diaries and scrapbooks to 1870, and home movies to 1928. Among the dozen members of my immediate family who have published, several count as historians, but several also as scientists, and I have worked extensively in both genres. In the 1980s I became interested in computational social science, and I programmed questionnaire writing software for two of my books, published in 1986 and 1989. This software offered the user several question formats, multiple response formats, and opportunities to write in text. For example, the user might want a series of agree-disagree items, with a particular number of responses (e.g., strongly agree, agree, neutral, disagree, strongly disagree – or leaving out neutral to force a positive or negative choice), and would write in a statement for each question to which the respondent would react. The software would save the questionnaire in a special format, display it so a person could answer the questions on the computer, then save the data for later analysis by another part of the software.
Simultaneously, I was working in the area of artificial intelligence called neural networks, publishing some neural net multi-agent system software as early as 1987. This got me thinking again about the field in which I had done my Harvard doctoral oral exam, ethnopsychiatry, the sociology and anthropology of mental disorder with an emphasis on the interplay of culture and personality. After I came to the National Science Foundation in 1992, the first workshop I funded was about artificial social intelligence, the application of machine intelligence techniques to social phenomena, including both theory building and data analysis. Shortly afterward I became involved in the Digital Library Initiative, just at the time the World Wide Web was being born, and I quickly saw the opportunity to use online questionnaires to develop culturally-rooted questions for archiving individual beliefs, attitudes, and values. In 1997, I launched a website called The Question Factory for this purpose, and a year later joined the team sponsored by the National Geographic Society that was planning the first of a pair of massive online questionnaires.
Individual human personalities are formed in a socio-cultural context, and societies vary, so many measures of personality must be culturogenic. That is, the questions must be drawn from the same socio-cultural context as the respondent, and cannot simply be invented by the researcher. For example, in Survey2000, the first major NGS online questionnaire, I included an open-ended item, asking people to predict what the world would be like in a century. I then went through the responses of something approaching 20,000 people, gleaning and collating statements about the future, resulting in 2,000 distinct statements. I then wrote software called The Year 2100 to administer these as questionnaire items. The respondent was asked to judge how good each idea was, on an 8-point scale, thereby measuring the person’s values in terms of the issues the surrounding culture thought would be important for the long-term future. Respondents also rated how likely each prediction was to come true on another 8-point scale. The correlation between these two scales represents the person’s optimism, which is believing good things will happen, normed in terms of the respondent’s own beliefs and values. The software permits analysis of optimism in each of 20 areas of life, and research I have done with it confirms that people may have complex patterns of optimism and pessimism across domains of experience.
Subsequently, I assembled materials from other cultural sources for nine more software modules, a total of 20,000 stimuli, each with 2 responses, for 40,000 measurements. Although standard psychology tests lack fine detail and are somewhat abstract, they are also useful for personality capture, so an eleventh module incorporates 2,000 public domain personality measures from the International Personality Item Pool (http://ipip.ori.org/) created by Lewis R. Goldberg (1999). Another module explores the individual’s short-term memory, inspired by the classic work by Saul Sternberg (1966), asking the individual to memorize 5,760 strings of digits or letters, measuring both accuracy and swiftness of recall. A new area of research is the emotional connotations of autobiographic memories (Bainbridge, 2006). Complete personality capture will also need to incorporate natural language processing voice recognition work on the person’s speech; motion capture of gestures, facial expressions, and task-oriented movements; and metabolic, anthropometric, and genetic measurements of the individual’s physical body. This sounds like a lot of effort, but with automation it could be finished within two or three months of the person’s effort, and might be scheduled as a free-time activity over several years.
Increasingly, we will archive ourselves as an incidental byproduct of interacting with our machines. For example, the Teachable Agents Group at Vanderbilt University has been exploring how to help students learn by having them teach science and math to a computer-animated artificial agent named Betty (Biswas et al., 2005). By recording the detailed interactions between Betty and the student, the system learns much about the student’s deep thought processes, precisely in a context where an artificial intelligence is learning to think like a person. A team at the University of Southern California has been recording the behavior of students in educational immersive game environments, in order to design better instructional games but incidentally recording much about the students (Yang et al., 2005). John Smart has argued that human-computer interaction technologies will continue to advance in the coming decades, and soon we will be using a linguistic user interface to converse with our machines:
As our own most-preferred digital personal interface (our “Digital Me”) gains exponentially more storage and processing capacity, it will incrementally engage in a process that William Sims Bainbridge calls “personality capture.” Our DM’s will carry an ever more valuable record of all the past communication we have had with them, and increasingly become our best professional representatives, coaches, managers, and extended memory for important events (Smart, 2005).
Rough calculations suggest that a medium to high fidelity copy of a single personality could constitute less than a gigabyte of data, perhaps ten gigabytes if this estimate is too low and the person’s genetic code and other physical data were included. The code to create the artificial intelligence personality emulation system, and the messages to tell extraterrestrials how to assemble the system, would add several more gigabytes. There would be economies of scale in sending information about many human beings, especially because many of the individual measurements are comparative with other humans. For example, it could be more efficient to send the genetic code as a list of the person’s 50,000 or so genes, rather than the 3 billion base pairs, but this assumes that a key to the genes has been sent separately. Fidelity of each avatar could increase as the number of people to be emulated increased, and as our artificial intelligence technology improved. These facts are quite compatible with interstellar transmission, because the recipient civilization would want time to study the information as it came in, and would probably learn much from low fidelity emulations of humans, increasing the fidelity as time passed.
Cosmopolis: The Emulation System
Many writers have speculated that already a number of civilizations are in radio communication with each other, and that we could join their cosmic club. I will now suggest that a club needs a clubhouse, and that each such civilization will establish a vast virtual reality world, perhaps on an unused asteroid or specially-built space station, call it Cosmopolis. Here, computer emulated representatives of each of the communicating civilizations would live and work together, serving to link the intelligent species of that solar system with the rest of the galaxy. Whether by local (i.e. interplanetary) radio or in person, members of the host civilization would interact with the avatars of other civilizations. But the hosts would also contribute their share of avatars to Cosmopolis. For many purposes, their AIs would need to be on an equal footing with the others.
Although contemporary radio technology could probably deliver information at a reasonable bit-rate across 100 parsecs, at this distance a single question and answer exchange would take 650 years. Because the Cosmopolis avatars would reside locally, within information systems equipped with sensors and robots, communication would be instantaneous, rather than delayed by the speed-of-light limitation on interstellar radio, and they would be able to act within the host civilization. They would function in realtime as ambassadors and as trading colonists providing translated versions of works of art, literature, and scientific discoveries.
When interacting with a biological being, a Cosmopolitan avatar would need the equivalent of a visual display and voice synthesis equipment. The technology for creating realistic avatars in collaborative virtual environments is already far advanced, and considerable research has examined the social implications of this technology (Bailenson, et al., 2004, 2005). The areas where research is not far enough advanced are natural language processing by computerized dialog systems, and the fundamental artificial intelligence technologies needed to emulate those parts of the human mind that cannot readily be captured in attitudes and preferences.
When the avatars are interacting with each other, and a biological being is not in the communication loop, there is no reason why the speed of thought and interaction could not be greatly accelerated. In discussing physical travel near the speed of light, we have become used to talking about time dilation. Inside Cosmopolis, we may encounter phenomena of time constriction. One possible paradox would be that the interacting cultures within the collaborative virtual environment would create fusion subcultures far more rapidly than the host civilization could assimilate their innovations. Thus, all of the participating civilizations could experience extreme cultural lag (Ogburn, 1922), once the avatars were communicating with each other effectively. Of course, this assumes they will be able to communicate effectively enough in the beginning to develop a culturally catalytic relationship.
In his early novel, Iceworld, Hal Clement suggested that the best way members of two very different intelligent species could begin to communicate would be by exchanging things of value. He understood that it might be difficult to predict what an extraterrestrial being would want. Technological inventions would be one possibility, but new technology depends so heavily upon existing industrial capacity, that only a few of our engineering triumphs might be valuable to extraterrestrials without requiring them to make a considerable investment in support infrastructure. Music is another possibility, and today’s synthesizer technology could readily transpose Bach to different octaves in sounds that would be sweet to alien ears. Perhaps we could sell large prime numbers, on the assumption that the other civilization could not possibly devise mathematical means to mass produce them. My point is that our avatar ambassadors must be prepared to offer many kinds of information that might be of value, at the same time they evaluate the worth to humans of what the ETs are offering.
This raises the question of who shall represent each world. The term ambassador implies a representative of a government. Given the direction Earth history is taking, where global capitalism vies for supremacy with religious fundamentalism, one just as well might argue the ambassadors would be salesmen or evangelists. If a civilization has no central government, will it have several competing ambassadors, representing whatever factions it possesses? The answer may be that Cosmopolis will be a trading colony, representing whatever individuals and groups were sufficiently motivated to send their avatars to the stars. We might well recall the opening comment from the Babylon 5 television series, about a place where humans and aliens could meet: “It’s a port of call, home away from home for diplomats, hustlers, entrepreneurs and wanderers.”
There are many good reasons to hope that well-established advanced civilizations will be able to develop cooperative relations with each other (Harrison, 1997), but the chances for conflict would seem greater the more fragmented each civilization is. Perhaps interstellar war is possible, after all, but it will take the form of some forces trying to prevent others from getting or giving information through Cosmopolis. The weaponry could consist of computer viruses infiltrating the alien equivalent of the World Wide Web. By placing the emulation system on an isolated asteroid, guarded by informatic firewalls, the host civilization would protect itself against these possible dangers.
Plan for Action
It might seem that any work on developing interstellar embassies would need to wait until after extraterrestrial civilizations had been contacted or plans were well-advanced for interstellar expeditions by means of highly advanced spacecraft. I think the opposite is true: The goal of direct contact with extraterrestrials via computer emulation could motivate the search for extraterrestrial intelligence.
Some individuals will want to invest energy and money in developing the needed personality capture and emulation technology because they personally hope to become interstellar ambassadors. If Cosmopolis would be more like a trading colony than a mere diplomatic office, then its virtual population could be quite substantial. I like to imagine it like the city of Hong Kong, trading between China and the world, which has a population of fully seven million. While most people might not be interested, it would seem reasonable to expect that more than seven million people alive today would be willing to invest significant sums, if the direction the research would take were well mapped out. If each of seven million people invested ten thousand dollars, the total would be $70,000,000,000. On an annual basis, even one thousandth of that could support the early stages of a vigorous research program.
Of course, many people would not believe that any conceivable technology could actually grant immortality through personality capture. Perhaps their religious views define immortality very differently, or they see compelling philosophical arguments why their avatars could not be themselves. When I published my factual article, “A Question of Immortality,” in Analog magazine, the title was chosen to reflect this controversy. The editor, Stanley Schmidt, insisted that I avoid claiming that personality capture was the same thing as real immortality, and therefore I ended the article with this paragraph:
We already possess technology that can allow aspects of your personality to influence the world in a dynamic manner, even after you are no longer living in it. This falls short of true immortality, because you (as conventionally defined) will not be conscious of it. Your evaluation of the possibilities for future cybernetic immortality will depend, therefore, not merely upon your estimate of the technical possibilities but also upon your personal conception of yourself. (Bainbridge, 2002a: 40)
Whether coincidentally or by design, the May 2002 issue of Analog that held my article also contained several stories about technological immortality, thereby illustrating the breadth of imagination that can be brought to bear on the topic. Sherry Turkle (1984) has argued that people who grow up from early childhood using a computer may come to feel it is a “second self.” Some people may feel that their avatars are more like descendents than a second self, what Hans Moravec (1988) called “mind children.” In the two decades since Turkle and Moravec wrote, a considerable literature has arisen in computer science about the nature of self-reference and the social-psychological relationship between people and their autonomous software agents or avatars (Smith, 1986; Friedman and Nissenbaum, 1997; Dobbyn and Stuart, 2003).
The crucial point is that many people, guided by a range of beliefs about themselves, will identify to a significant extent with their avatars. Of these, a small percentage but large absolute number (given the 6,000,000,000 population of our planet) could be enlisted in a social movement to send virtual ambassadors to the stars. Logically, they should invest their efforts along two parallel tracks: (1) developing the cognitive and information technology needed for high-fidelity emulation of personality, and (2) locating extraterrestrial civilizations while building the infrastructure to communicate with them. Both of these tracks offer ample opportunity for individuals and small groups to contribute.
As proponents of the search for extraterrestrial intelligence have found from bitter experience, governments are not reliable sources of funding for anything that requires imagination and an orientation toward the future. Governments do support some research on artificial intelligence, human-computer interaction, and information archiving technologies. However, government support for fundamental cognitive science is weak, and none of the research grants in that area are likely to go for personality capture or emulation. Thus, individuals and groups should identify research gaps that realistically could become the target of their own intellectual efforts. At the same time, leaders of the SETI movement and well-meaning entrepreneurs should be alert to opportunities to channel money and volunteer labor into projects that could have near-term success goals as well as aiming toward direct contact with extraterrestrials in the more distant future.
The broad plan described here could add motivation to invest in completing the instruments currently under construction or design to detect extraterrestrial radio signals. A logical next step is to beam radio messages outward, containing personality-capture information from individual investors, even though we do not yet know the most likely targets. Sending information for an avatar of oneself to the stars could have a symbolic benefit, very compelling for some wealthy visionaries, even if the realistic chances of the message being received are low. Some of their investment could go to search efforts, with the understanding that as soon as an extraterrestrial civilization were found, the messages would be resent in its precise direction.
If the next generation or two of attempts to detect extraterrestrial signals were unsuccessful, the campaign would merely move to a higher level. The efforts of amateurs, guided by a few scientists in the appropriate fields, could develop effective techniques for personality capture and emulation. People who wanted their personalities sent to the stars could assign investments over to special non-profit organizations that served as terrestrial personality archives and extraterrestrial development promoters. Once sufficient resources were concentrated, and space technology had advanced somewhat further, we could build our own Cosmopolis on the Moon.
Inside a functional lunar Cosmopolis, cognitive scientists and computer scientists could complete development of the needed Starbase technology. Either ordinary humans living in a lunar outpost, or the artificial intelligence agents dwelling in Cosmopolis themselves, could use teleoperation robots to build an exceedingly large SETI antenna array on the shielded far side of the Moon. Lunar and orbiting telescopes should, by that time, be able to identify nearby extrasolar planets capable of supporting life. If the lunar radio detection effort failed, there would be no sense of defeat, but a redirection of efforts toward construction of a fleet of interstellar vehicles to send multiple Starbases on colonizing missions and in tireless search for other civilizations.
References
Bailenson, Jeremy N., Andrew C. Beall, Jack Loomis, James Blascovich, and Matthew Turk. 2004. “Transformed Social Interaction: Decoupling Representation from Behavior and Form in Collaborative Virtual Environments.” PRESENCE: Teleoperators and Virtual Environments, 13(4), 428-441.
Bailenson, Jeremy N., Andrew C. Beall, James Blascovich, Jack Loomis, and Matthew Turk. 2005. “Transformed Social Interaction, Augmented Gaze, and Social Influence in Immersive Virtual Environments.” Human Communication Research, 31, 511-537.
Bainbridge, William Sims. 1986. Experiments in Psychology. Belmont, California: Wadsworth.
Bainbridge, William Sims. 1987. Sociology Laboratory. Belmont, California: Wadsworth.
Bainbridge, William Sims. 1989. Survey Research: A Computer-Assisted Introduction. Belmont, California: Wadsworth.
Bainbridge, William Sims. 1993. “New Religions, Science and Secularization,” pp. 277-292 in The Handbook of Cults and Sects in America, edited by David G. Bromley and Jeffrey K. Hadden. Greenwich, Connecticut: JAI.
Bainbridge, William Sims. 1994. “Extraterrestrial Intelligence: Communication.” Pp. 1200-1203 in The Encyclopedia of Language and Linguistics, edited by R. E. Asher. Oxford: Pergamon.
Bainbridge, William Sims. 2002a. “A Question of Immortality,” Analog, 122(5): 40-49.
Bainbridge, William Sims. 2002b. “The Spaceflight Revolution Revisited,” pp. 39-64 in Looking Backward, Looking Forward, edited by Stephen J. Garber. Washington, D.C.: National Aeronautics and Space Administration.
Bainbridge, William Sims. 2003 “Massive Questionnaires for Personality Capture,” Social Science Computer Review 21 (3): 267-280.
Bainbridge, William Sims. 2004. “The Future of the Internet: Cultural and Individual Conceptions.” Pp. 307-324 in Society Online: The Internet in Context, edited by Philip N. Howard and Steve Jones. Thousand Oaks, California: Sage.
Bainbridge, William Sims. 2006. “Cognitive Technologies.” Pp. 207-230 in Managing Nano-Bio-Info-Cogno Innovations: Converging Technologies in Society, edited by William Sims Bainbridge and Mihail Roco. Berlin: Springer.
Bainbridge, William Sims, Edward E. Brent, Kathleen Carley, David R. Heise, Michael W. Macy, Barry Markovsky, and John Skvoretz. 1994. “Artificial Social Intelligence,” Annual Review of Sociology 20: 407-436.
Biswas, Gautam, Krittaya Leelawong, Dan Schwartz, Nancy Vye, and the Teachable Agents Group at Vanderbilt. 2005. “Learning By Teaching: A New Agent Paradigm for Educational Software,”Applied Artificial Intelligence, vol. 19, (pp. 363-392).
Clement, Hal. 1953. Iceworld. New York: Gnome Press.
Dick, Steven J. 2003. “Cultural Evolution, the Postbiological Universe and SETI,” International Journal of Astrobiology, 2(1): 65-74.
Dobbyn, Chris and Susan Stuart. 2003. “The Self as an Embedded Agent,” Minds and Machines, 13(2): 187-201.
Friedman, Batya, and Helen Nissenbaum. 1997. “Software Agents and User Autonomy.” Pp. 466-469 in Proceedings of the First International Conference on Autonomous Agents. New York:Association for Computing Machinery.
Goldberg, Lewis R. 1999. “A Broad-bandwidth, Public Domain, Personality Inventory Measuring the Lower-level Facets of Several Five-factor Models.” Pp. 7-28 in Personality Psychology in Europe, edited by I. Mervielde, I. Deary, F. De Fruyt, and F. Ostendorf , volume 7. Tilburg, The Netherlands: Tilburg University Press.
Harrison, Albert A. 1997. After Contact: The Human Response to Extrtaterrestrial Life. New York: Plenum.
Morovec, Hans. 1988. Mind Children: The Future of Robot and Human Intelligence. Cambridge, Massachusetts: Harvard University Press.
Ogburn, William Fielding. 1922. Social Change with Respect to Culture and Original Nature. New York: B. W. Huebsch.
Robinett, Warren. 2003. “The Consequences of Fully Understanding the Brain,” pp. 166-170 in Converging Technologies for Improving Human Performance, edited by Mihail Roco and William Sims Bainbridge. Dordrecht, Netherlands: Kluwer.
Smith, Brian Cantwell. 1986. “Varieties of Self-Reference.” Pp. 19-43 in Proceedings of the 1986 Conference on Theoretical Aspects of Reasoning about Knowledge. New York: Association for Computing Machinery.
Sternberg, Saul. 1966. “High-Speed Scanning in Human Memory,” Science 153: 652-654.
Turkle, Sherry. 1984. The Second Self: Computers and the Human Spirit. New York: Simon and Schuster.
Yang, Kiyoung, Tim Marsh, Minyoung Mun, and Cyrus Shahabi. 2005. “Continuous Archival and Analysis of User Data in Virtual and Immersive Game Environments,” Proceedings of CARPE ‘05, pp. 13-22.
Comments (0)
You don't have permission to comment on this page.