% The bibliography style file is called spiebib.bst, % which replaces the normal LaTeX style unstr.bst. % One departure from the specifications found in unstr.bst % is the addition of the `journal' field to the \inproceedings % entry type, whose use is demonstrated in Ref. 5 (Hanson93c). %% \documentstyle[spie]{article} \documentclass[]{article} % use this for latex2e \usepackage{spie} % use this for latex2e %>>>> psfig.sty to include EPS figures; comment out if not needed % \input{psfig} \def\arraystretch{1.2} % make table vertial spacing slight larger \makeatletter %%%% debut macro %%%% %% ---------------------------------------------------- %% Copyright (c) 1993 Hydro-Quebec mboyer@robot.ireq.ca %% ---------------------------------------------------- %% Bring items closer together in list environments % Prevent infinite loops \let\orig@Itemize =\itemize \let\orig@Enumerate =\enumerate \let\orig@Description =\description % Zero the vertical spacing parameters \def\Nospacing{\itemsep=11pt\topsep=0pt\partopsep=0pt\parskip=-8pt\parsep=0pt} \def\noitemsep{ % Redefine the environments in terms of the original values \renewenvironment{itemize}{\orig@Itemize\Nospacing}{\endlist} \renewenvironment{enumerate}{\orig@Enumerate\Nospacing}{\endlist} \renewenvironment{description}{\orig@Description\Nospacing}% {\endlist} } \def\doitemsep{ % Redefine the environments to the original values \renewenvironment{itemize}{\orig@Itemize}{\endlist} \renewenvironment{enumerate}{\orig@Enumerate}{\endlist} \renewenvironment{description}{\orig@Description}{\endlist} } %%%% fin macro %%%% \makeatother %% Use \noitemsep to get small list spacing %% Use \doitemsep to get large list spacing \title{Life at the Limits of Physical Laws} %>>>> The author is responsible for formatting the % author list and their institutions. Use \skiplinehalf % to separate author list from addresses and between each address. % The correspondence between each author and his/her address can be % indicated with a superscript in italics, % which is easily obtained with \supit{}. \author{Robert J. Bradbury \skiplinehalf Aeiveos Corporation, Seattle, WA, U.S.A. \skiplinehalf Copyright \copyright \hspace{0.1mm} 2001, Robert J. Bradbury } %>>>> Further information about the authors, other than their % institution and addresses, should be included as a footnote, % which is facilitated by the \authorinfo{} command. \authorinfo{Further author information: E-mail: bradbury@aeiveos.com} %% NB: when using amstex, you need to use @@ instead of @ %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %>>>> uncomment following for page numbers %% \pagestyle{plain} %>>>> uncomment following to start page numbering at 301 \setcounter{page}{301} \begin{document} \maketitle %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \begin{abstract} Some of the problems that plague SETI research are the problems of the abundance of liquid water planets, the probability of the development of intelligent life, whether or not intelligent life forms develop technology, how long intelligent technological civilizations may survive, and whether or not interstellar travel or colonization are feasible or affordable. These problems lead to extensive and potentially irresolvable debates regarding the various paths species and civilizations may follow from a primitive level to our current level and beyond. This discussion will focus instead on the question of what the characteristics of intelligent technological life should be at the limits of known physical laws. Why do this? Well, because as Scotty observed on the Starship Enterprise, ``Captain, I canna change the laws of physics!''. Even if the laws of physics do not deny the feasibility of a life form, the lack of a practical engineering path to it may prevent its existence. At these limits, the form(s) that life takes may be clearer because convergent evolution could drive civilizations into a very limited set of ecological niches. An architecture for civilizations that hits many of these limits will be proposed. Its characteristics include thought capacities in excess of a trillion trillion times that of an individual human, survival times of trillions of years and astronomical observational capacities trillions of times greater than our civilization. Such civilizations, should, over time, become the dominant population of galaxies. Our own civilization may reach this state within this century. The impact of these conclusions on classical radio and optical SETI verses astrometric and occultation astronomy will be discussed. \end{abstract} %>>>> Please include a list of keywords after the abstract \keywords{Dyson shells, evolution, Matrioshka Brains, megascale engineering, nanotechnology, optical SETI, radio SETI, technological civilizations} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \section{INTRODUCTION} \label{sect:intro} % \label{} allows reference to this section For the past 40 years SETI proponents have searched for extraterrestrial civilizations unsuccessfully. Various explanations have been proposed for this. These include: \noitemsep \begin{itemize} \item the number of stars that have been searched is small; \item the \emph{Rare Earth} hypothesis\cite{PDW2000} limiting the abundance of intelligent life; \item the Earth is under quarantine \cite{JAB1973}; being ``cultivated'' for information \cite{DGS1982}; or is subject to interdiction \cite{MJF1987} \end{itemize} \doitemsep The lack of success is causing more scientists to suggest that in fact we may be alone \cite{IC2000}, that we are the first technological civilization to evolve in our galaxy. This paper adopts the perspective that the glass may be much more full than empty, i.e. that advanced technological civilizations may be quite abundant. In the Milky Way, 10 billion stars may have sufficient metallicity to produce terrestrial planets older than 5 billion years\cite{VT1982}. Recent estimates suggest more than 70\% of the Earth's in the galaxy should be older than ours\cite{CHL2000}. If the characteristics that make our Earth \emph{special}, such as a comet intercepting Jupiter and a large moon\cite{PDW2000}, are not \emph{too} rare \emph{or} hazardous environments actually accelerate the rate of the evolutionary development of intelligence, then we should expect most technological civilizations to be far more advanced than we ourselves currently are. Here, we will propose models for advanced civilizations and explore why their capabilities can explain the apparent lack of signals. The reasons the prospects for \emph{communications} with extraterrestrial intelligence (CETI) are dismal will be discussed. Finally, the contributions that emerging branches of astronomy can make to programs that \emph{search} for extraterrestrial intelligence (SETI) will be explored. \section{NANOTECHNOLOGY} \label{sect:Nanotechnology} Although the concept of nanotechnology was developed at about the same time as CETI in 1959,\cite{RPF1961,GC1959} it received little attention until 1986 \cite{KED1987}. It has only been developed into a robust branch of science within the last decade \cite{KED1992,RAF1999}. While nanotechnology has had its share of detractors, it is increasingly becoming recognized as an important area of study, one that is intimately intertwined with organic chemistry, materials science, biotechnology, computer science, nanoscale materials, micromechanical and electrical systems (MEMS) and molecular electronics. Almost every issue of the prestigious journals {\it Science} and {\it Nature} now contain articles involving nanoscale engineering or enabling scientific developments. Technological civilizations, as they develop, expand their capabilities to both the smaller and larger scales. Extrapolations from the current microelectronic industry lithography scale of 0.18 $\mu$m, along the trend predicted by Moore's Law, leads to a rather hard limit of atomic scale manufacturing around 2040. Molecular nanotechnology has a number of important features. These include atomically precise assembly, self-assembly, self-replication, self-motility and the execution of resource acquisition, survival and reproduction programs. Single-celled biological organisms possess all of these properties and constitute an existence proof that molecular nanotechnology is feasible. The primary differences between the naturally evolved nanotechnology found in biological systems and more robust forms envisioned in the previously mentioned references is the use of solution chemistry (so the precise position of all atoms is not always known) and materials that have strengths and operating temperature ranges that are less than the theoretical limits. A logical progression exists for the development of molecular nanotechnology based on alternate forms of carbon (fullerenes and buckytubes), followed by diamond, sapphire and titanium carbide. Progress along these and related paths eventually leads technological civilizations to develop an extremely large range of devices and structures that are precisely manufactured from individual atoms. Molecular nanotechnology enables the magical ``replicators'' of the type seen in Star Trek. Anything that can exist should be able to be assembled. It does have limits however. Heat removal constraints will limit the production rate of assemblers operating at the atomic scale.\cite{KED1992} And nanocomputer throughput will be limited by the amount of entropy generating bit-erasure that is done that must removed as heat.\cite{RL1961} \section{SELF-REPLICATING SYSTEMS} \label{sect:SelfReplicatingSystems} Some of the first work on the feasibility of artificial self-replicating systems was done by John von Neumann in the 1940s, though it was not published until the mid-1960s \cite{JVN1966}. NASA later studied manufacturing systems that could replicate themselves for space colony development\cite{RAF:1982}. Though this study appears to have been forgotten by many in the aerospace industry, it is widely known among nanotechnology researchers. Biological self-replicating systems (e.g. bacteria) have mass doubling times as short as 20 minutes \cite{TDB1991,JCS1990}. These are certainly sub-optimal as these genomes contain several thousand genes while the minimal genome for self-replication may only require $\sim350$ genes \cite{ARM1996}. We now have sequenced the genomes of dozens of these microorganisms and as a better understanding of the function of all of the genes they contain develops, the design requirements for robust self-replicating systems will be made clear. The limits for the mass doubling times are less clear. Estimates as short as 2 milliseconds using assembly line methods have been postulated \cite{JSH1999}, though these rates are likely to be constrained by heat removal requirements. The difference between the energy levels of high energy state intermediates and the final bonded energy levels determines the amount of waste heat produced in chemical reactions. Initial designs for self-replicating systems are likely to be less efficient and therefore constrained to operate more slowly than systems that are developed over time that utilize optimized reactions that minimize heat production. \section{THE DEVELOPMENT OF ADVANCED TECHNOLOGICAL CIVILIZATIONS} \label{sect:StagesOfDevelopment} A Kardashev Type-I planetary civilization \cite{NSK1964} (KT-I) reaching the limits of resources on its planet should begin the transition to a Type-II stellar civilization (KT-II). Freeman Dyson first suggested this over 40 years ago \cite{FJD1960,FJD1960b}. He pointed out that at the low growth rate of 1\%, civilizations would require all of power produced by their star in only 3000 years. If the growth rates in power consumption seen by our civilization over the last 200 years continue, the time for us will be closer to 1300 years. With the development of space stations relying on solar power, humanity has begun this transition. \subsection{Dyson Shells} Unfortunately Dyson limited his vision to the construction of ``biospheres'' around stars. This has resulted in misdirected searches for Dyson shells since that time.\cite{RJB2001a} Even if a civilization utilizes most of its planetary matter to construct biospheres, it is still likely to want to lay claim to all of the energy being produced by the star. This is simply because there are a number of very useful things that can be done with that amount of energy. Little thought seems to have been devoted to what applications might be enabled by civilizations with $\sim10^{26}$ W at their disposal. Some the author has considered include: \noitemsep \begin{itemize} \item Interstellar probe or ship propulsion: \begin{itemize} \item 1\% allows the annual launch of 10 billion Daedalus flyby probes\cite{PD1978} to nearby stars. \item 10\% allows the annual launch of 1 billion personal starships (Daedalus type ships with sufficient fuel to decelerate in the target system). \end{itemize} \item Interstellar communication: \begin{itemize} \item 1\% allows 10 billion 100 TW laser beams \item 10\% allows circulating photonic memory of $10^{22}$ TBytes/sec \end{itemize} \item 10\% allows reconfiguring of all the atomic bonds in a solar system in $\sim500$ years \item Long term navigation of entire solar systems to avoid hazards such as gamma-ray bursts, supernovas, collisions with wandering planets, etc. \end{itemize} \doitemsep \subsection{Matrioshka Brains} Using self-replicating nanomachinery, it is feasible to consider disassembling the planets to harvest all of the solar energy available. This is merely our current satellite and space station construction activities taken to their logical limits. The following table contains estimates of the usable material in various solar system objects and the results of some simulations of planetary disassembly. The areal density of the solar arrays used for computing these disassembly times was 1 kg/m$^2$, which is slightly better than best available arrays we now manufacture. It was also assumed that the collectors remained in orbits with radii similar to that of the planet. \begin{center} \begin{tabular}{|l|c|c|c|c|r@{.}l|} \hline & \multicolumn{1}{|c|}{\bf Useful} & \multicolumn{1}{|c|}{\bf Orbital} & \multicolumn{1}{|c|}{\bf Disassembly} & \multicolumn{1}{|c|}{\bf Exponential} & \multicolumn{2}{|c|}{\bf Disassembled} \\ \multicolumn{1}{|c|}{\bf Body} & \multicolumn{1}{|c|}{\bf Mass} & \multicolumn{1}{|c|}{\bf Radius} & \multicolumn{1}{|c|}{\bf with $10^{26}$ W} & \multicolumn{1}{|c|}{\bf Self-disassembly} & \multicolumn{2}{|c|}{\bf Areal Density} \\ \hline \multicolumn{1}{|c|}{} & \multicolumn{1}{|c|}{(kg)} & \multicolumn{1}{|c|}{(km)} & \multicolumn{1}{|c|}{time} & \multicolumn{1}{|c|}{time} & \multicolumn{2}{|c|}{kg/m$^2$} \\ \hline Mercury & $3.3\times10^{23}$ & $5.8\times10^{7}$ & 5 hours & 14 days & 7&8 \\ \hline Venus & $4.9\times10^{24}$ & $1.1\times10^{8}$ & 16 days & 114 days & 33&1 \\ \hline Earth & $5.9\times10^{24}$ & $1.5\times10^{7}$ & 22 days & 179 days & 20&9 \\ \hline Mars & $6.4\times10^{23}$ & $2.3\times10^{8}$ & 12 hours & 176 days & 1&0 \\ \hline Jupiter & $1.9\times10^{27}$ & $7.8\times10^{8}$ & 563 years & 691 years & 9&6 \\ \hline Saturn & $5.7\times10^{26}$ & $1.4\times10^{9}$ & 60 years & 181 years & 3&4 \\ \hline Uranus & $8.7\times10^{25}$ & $2.9\times10^{9}$ & 3.3 years & 223 years & 0&7 \\ \hline Neptune & $1.0\times10^{26}$ & $4.5\times10^{9}$ & 8.2 years & 624 years & 0&32 \\ \hline Pluto & $1.3\times10^{22}$ & $5.0\times10^{9}$ & 2 minutes & 266 years & 0&000037 \\ \hline Asteroids & $5.9\times10^{21}$ & $4.1\times10^{8}$ & very fast & very fast & 0&0025 \\ \hline \end{tabular} \\ \emph{These numbers should be used for relative comparison purposes only. \\ They are subject to change as the simulation models are improved.} \end{center} The exponential growth allowed by self-replicating nanomachinery and the increasing amount of power that can be delivered back to the planetary body as the solar array grows, allow relatively short disassembly times.\footnote{The actual limits depend in part on how thin the solar collectors might be made. The theoretical limits\cite{KED1992b,FJD1995} are around $10^{-3}$ kg/m$^2$ and might be approached with space-based manufacturing methods. There is a problem that towards the end of the dismantlement process the heat capacity of the planet is exceeded. If creative solutions for heat removal from the planet are developed, disassembly times could be even faster. A more detailed discussion of these issues may be found in \emph{Planet Disassembly}\cite{RJB1998}.} Once a fully functional star enshrouding collector array is built the other planets can be disassembled at an even faster rate because the full power of the star is available. One conclusion that may be drawn from this is that within our solar system, almost any major body provides sufficient material to enshroud the sun. Even the asteroids may be sufficient if the collectors are moved somewhat closer to the sun. So it seems likely that in most solar systems in which technological civilizations find themselves they will have the resources to harvest the power output of their star. Another conclusion is that when a civilization ``decides'' to make the KT-I (planetary) to KT-II (stellar) transition it \emph{may} do it \emph{very} quickly. Once the material found in solar systems is lifted out of its respective gravity wells, a long process of relocating it into optimal orbits is required. This may involve moving carbon (derived from methane), oxygen (derived from ice) and nitrogen (derived from ammonia) into the inner regions of the solar system and moving metals with desirable magnetic or superconducting properties to the outer regions of the solar system. As this is done, increasing amounts of power are dedicated to the construction and operation of nanocomputers associated with the power collecting satellites orbiting the star. The nanocomputers are interlinked with one another via arrays of VCSEL lasers and CCD detectors. Over time a multi-layer Dyson shell architecture arises where each layer operates at a specific temperature that is related to the elements available and computer architectures chosen to perform calculations. These architectures may range from high-temperature mechanical-logic computers near the star to very cool computers utilizing superconductor based logic at distances light-hours from the star. This solar system scale supercomputer manufactured from atomic scale parts has been named by the author a ``Matrioshka Brain'' (after the nesting Russian dolls). The next table compares the computational and memory capacities of a human and humanity with a single node of a Matrioshka Brain and the entire Matrioshka Brain. \begin{center} \begin{tabular}{|l|c|c|c|c|} \hline & \multicolumn{1}{|c|}{\bf Human} & \multicolumn{1}{|c|}{} & \multicolumn{2}{|c|}{\bf Matrioshka Brain } \\ \cline{4-5} {\bf Characteristic} & {\bf Brain} & {\bf Humanity} & {\bf Node} & {\bf Solar System} \\ \hline Operations / second & $10^{14} - 10^{17}$ & $10^{24} - 10^{27}$ & $> 10^{21}$ & $> 10^{42}$ \\ \hline Bits of memory & $10^{9} - 10^{20}$ & $10^{19} - 10^{30}$ & $> 10^{24} $ & $> 10^{45} $ \\ \hline \end{tabular} \end{center} It can be seen that there are rather large scale differences. How does this impact the probability of CETI? A couple of examples of the gap between a Matrioshka Brain at the Kardashev Type-II civilization level and our pre-Type-I civilization level may be useful: \begin{itemize} \setlength{\itemsep}{-0.5mm} \item{The thought capacity of a Matrioshka Brain is so large that in a single microsecond it exceeds the combined thought capacity of all \emph{Homo sapiens} that have ever lived.} \item{The difference between the neuron count of a human and a nematode is $\sim10^8$. The difference between the thought capacity of a Matrioshka Brain and a humanity is $> 10^{15}$.} \end{itemize} Since the combined thoughts of humanity are \emph{puny} relative to the capacity of a Matrioshka Brain and we don't really ``communicate'' with nematodes, it seems very doubtful that these advanced entitites would have any desire to communicate with us! They may also be a fundamentally different type of intelligence. Between humans, our communication bandwidth is very asymmetric. Our primary input is visual and its capacity may be several megabytes per second. Our outputs, such as speech or typing, use completely different mediums and are limited to around a few dozen bytes per second. In contrast, the innermost internode communication capacities in a Matrioshka Brain are tens to millions of terabytes per second and they may have equivalent input and output bandwidths. So the individual "nodes" are likely to function as a more tightly integrated mega-mind than humanity, in its current form, ever could. Further details of this are discussed in papers at the \emph{Matrioshka Brain Home Page}\cite{RJBMBHP}. \subsection{Physical \& Engineering Limits} \label{sect:PhysicalLimits} The theoretical limits of megascale computing have been discussed by Moravec\cite{HPM1987b}, Sandberg\cite{AS1999} and Lloyd\cite{SL2000}. Reaching these theoretical limits requires sub-atomic scale engineering abilities -- something that may not even be feasible. If it is feasible, ``the step from normal matter is likely to be big and difficult'' as Moravec points out. So we can expect that extraterrestrial civilizations are likely to reach a developmental plateau at atomic scale engineering. Previous papers\cite{RJB1999,RJB2000} discuss the megascale engineering limits in more detail, particularly with regard to Matrioshka Brains. Here we will simply summarize some conclusions that have been reached. \noitemsep \begin{itemize} %% \\setlength{\itemsep}{-0.5mm} \item{There is insufficient material in the solar system, separate from the sun, to optimally utilize the energy it produces.} \item{The element abundances are unlikely to be optimal for designing the best computing architectures. So massive breeder reactors will be required to address this situation over the long term.} \item{Computing architectures consisting of all the non-stellar material in the solar system can be completely disassembled and reassembled in decades once the material is extracted from the gravity wells (if all of the solar energy is dedicated to this purpose).} \item{Cooling requirements set a hard limit computational density.} \item{The speed of light sets a hard limit on computational throughput.} \item{A small fraction of the material of the solar system provides KT-II level civilizations with billions of lunar diameter telescopes. Because these can be arrayed as an interferometer, their observational capabilities are rather staggering.} \item{They would be able to launch billions of small interstellar probes or ships or matter data "packets" annually.} \end{itemize} \doitemsep Communication between Matrioshka Brains is a formidable challenge. Even using a 40 GHz signaling rate, the limit of current semiconductor laser technology, billions of years would be required to transmit 1 trillionth of its knowledge base. If the entities were in relatively close proximity to each other, and large arrays of lasers are used, it could be feasible to exchange a not-insignificant fraction of your knowledge base using a sophisticated symbolic communications shorthand. It is doubtful that we could interpret this even if the entities were so careless as to allow communication photons to leak off into space. Unless methods like this are used, it is doubtful that anyone ever gets past the first page of \emph{Encyclopedia Galactica}. \subsection{Mind Uploading} \label{sect:Uplaoding} The author is aware of four respected scientists who are of the opinion that ``mind uploading'', the transfer of mental thinking capacity and consciousness from the human brain into a computer, is likely to be feasible. These include Marvin Minsky, Hans Moravec, Ray Kurzweil and W. Daniel Hillis. Two of these individuals have explored how this might be accomplished at the physical level.\cite{HPM1987,RK1999} This possibility means that we can foresee a logical progression in the development of technological civilizations. First they develop to the point of understanding the laws of physics and chemistry. Then they comprehend their basic biology and their information carrier (e.g. DNA). Then they develop computing machinery and a means for transferring their minds into those devices. If they choose to distribute their intelligence over a large enough physical volume (planetary volumes for example) or simply make backup copies of their intelligence (on opposite sides of a solar system), they have effectively immortalized themselves because local accidents cannot destroy their ``minds''. Over time, accidents will eliminate individuals and civilizations that do not follow this path. We can thus predict that that civilizations that do follow this path, \emph{should}, at some point, become the dominant populations in galaxies. \section{CETI vs. SETI} Many previous CETI projects have assumed that extraterrestrial civilizations of interest to us, should produce electromagnetic radiation, either by accident or intent, that we could detect. Those signals produced by accident would be similar to those that we ourselves currently produce. The low power levels of these signals makes them difficult to detect at moderate interstellar distances. In addition, terrestrial interference conspires against radio searches for signals similar to our own. Leakage signals would presumably decrease as a civilization matures. As a civilization develops more optimized means of transmitting information such as microwave, local lasers, fiber optics, low power multi-hop radio, etc. wasteful practices such as high power broadcast signals should diminish. Because leakage signals come from civilizations around our level, and are not designed to be received and understood, they are presumably of less interest than intentional communications from more advanced civilizations. Most radio CETI searches to date have been conducted with the assumption that an advanced civilization would be transmitting at frequencies at which others might be listening and that they are advanced enough and/or kind enough to continue transmitting for year after year until we hear them. This cost of this effort is usually justified with some hand-waving that we \emph{have} transmitted signals and we would want to do it continuously if we could afford it. But would we? If we had a choice of transmitting to a more advanced civilization or a less advanced civilization which would we choose? Our current CETI strategy seems to suggest we are looking for a Galactic Club handout. Situations where we communicate with those less advanced than us are cases where the difference is not that great -- probably only a few orders of magnitude for human infants and such animals as chimpanzees, dogs and cats. What rationale is there for extraterrestrials many orders of magnitude higher on the intelligence scale behaving any differently towards us than we do towards most of the other species on the planet? Due to the lack of success in CETI at radio frequencies and the possible advantages offered by the visible frequencies, CETI searches are now conducted in both regions. However both radio and optical CETI operate by pointing their telescopes at \emph{visible} stars! The civilizations that would seem best able to afford the power costs of interstellar communication would be those who have the greatest amount of power at their disposal, i.e. those that have enshrouded their stars with power harvesting devices. This process causes the star to grow dim or even disappear (at visible wavelengths). The lack of success in previous CETI programs is not surprising because the stars being examined most likely contain no civilizations or civilizations below our level of development. Civilizations that develop much beyond our level should make the KT-I to KT-II transition. After that they have all the power output of a star at their disposal. Using their telescopes they can see any civilizations with whom they might want to communicate. Civilizations that value their time and energy will attempt to communicate with more intelligent civilizations or civilizations with access to information that is otherwise unavailable to them. There seems to be little justification for directing time or energy towards civilizations that are lower than ``worms'' from their viewpoint. The phase space of devices that may be constructed using nanotechnology on solar system scales and the variety of underlying computing architectures that may be adopted by advanced civilizations may produce a strong motivation for leaving developing civilizations untainted to promote galactic ``diversity''. The development of a civilization along its own path is likely to foster such diversity because as the saying goes, ``Necessity is the mother of invention''. Premature exposure of developing civilizations to advanced extraterrestrial technologies could drive such civilizations onto previously explored development paths resulting in a reduction of the exploration of the phase space. Thus the Zoo Hypothesis\cite{JAB1973} may be correct if a Matrioshka Brain is using our planet as the subject of a controlled experiment. Alternatively, if the extraterrestrials desire to maximize the information content of the universe\cite{DGS1982} then the Interdict Hypothesis\cite{MJF1987} is the likely explanation for the lack of signals or extraterrestrial artifacts. These lines of reasoning argue that CETI at our level of development is unlikely to succeed. If that is the case what are the prospects for SETI? \subsection{Gravitational Microlensing} Astronomers have failed to account for the missing baryonic dark matter for many years. The gravitational microlensing surveys such as the MACHO project \cite{CA2000:ApJ542,CA2000:ApJ734} are finding objects that cannot to be attributed to brown dwarfs or white dwarfs \cite{DBF2000:ApJ265}. A Matrioshka Brain built around a central star, will still cause a gravitational microlensing effect. When traditional astronomical explanations for these objects are failing to produce results, it may be time to consider artificially engineered objects as possible sources. Due to the problems of the short lifetime of larger stars and the large material requirements for the power collectors and radiators of the outer shells of Matrioshka Brains there may be an optimal mass range for Matrioshka Brains that is somewhat smaller than the sun, yet larger than a brown dwarf. This concept is consistent with the MACHO observations. Efforts to combine the microlensing observations with IR and occultation astronomy (discussed below) could yield valuable information with regard to what the microlensing objects in really are. \subsection{Infrared Astronomy} It was Dyson who observed that sun enshrouding advanced civilizations can be observed by their infrared heat signature\cite{FJD1960}. He thought such civilizations have radiation peak emissions in the near IR around 10 microns\cite{FJD1960}, consistent with a liquid-water based civilization. If the civilization relocates itself into nanocomputers its possible range of operating temperatures becomes much greater. Marvin Minsky, in a discussion with Dyson,\cite{FJD1973} and later Suffern\cite{KGS1977} both observed that the most efficient energy utilization strategies from a thermodynamic standpoint would be to radiate the waste heat at temperatures slightly above the background radiation temperature. This will make them very difficult to observe directly. Civilizations must go through a development stage before they reach this very cool state. In solar systems where civilizations choose to dedicate much of the early material harvested to computer construction rather than power collection, in systems that are metal resource poor, and in systems where large amounts of power must be dedicated to material relocation to optimum temperature ranges -- the process of the star being enshrouded may not occur quickly. It may take decades, centuries or even millennia. In these systems we will observe the star slowly growing darker and darker, radiating increasingly more amounts of its radiation in the infrared, until finally it ``disappears''. We can use measurements from astrometric and other long term survey missions to identify stars with these characteristics and analyze their IR emissions for this type of process. The final state of the system will depend on the amount of construction material available. It does not appear that we could currently fill the outermost shells of a Matrioshka Brain and we would end up radiating at a temperature below that of liquid nitrogen, but above that of the microwave background radiation. So there is a possibility of SETI catching civilizations when they are making the KT-I to KT-II transition. Stars possessing the slowly varying brightness characteristic are known as Long-Period Variable stars. The General Catalog of Variable Stars\cite{GSVC1997} lists thousands of stars in these categories and more are being discovered every day. Astronomers who study these stars have recently reported that of the stars in this group, there are apparently more stars getting dimmer than are getting brighter\cite{JAM1998}. While this may be a characteristic of the stars in this class, these stars should examined in greater detail, particularly at infrared wavelengths, for signs of astroengineering activity. As the number of stars observed by current and planned surveys increases from the millions to the billions, we will begin to accumulate the data necessary to determine the frequency at which this transition occurs and therefore begin to set limits on the abundance of civilizations slightly ahead of our own level in the galaxy. \subsection{Occultation Astronomy} Dyson has proposed the value of occultation astronomy in the hunt for comets and planets\cite{FJD1990,FJD1992}. At the time his proposals were made, the technology to implement them would have been very expensive. That is not true today. Dyson did not consider the possibility of dark objects traveling through the galaxy that had the size of solar systems. Nor did he consider whether KT-II level civilizations might build billions of telescopes the diameter of the moon (or a much greater number of telescopes of more modest proportions). It appears clear that if advanced civilizations do exist and expand their capabilities in the ways described in this paper that occultation astronomy may be an excellent way to conduct SETI. Ultimately a combination of the three approaches -- gravitational microlensing, infrared and occultation may be needed to provide concrete identification of advanced technological civilizations. Highly evolved individuals of advanced technological civilizations may decide to sever their connection with their ``parent'' Matrioshka Brain in which they ``live'', even though this would require sacrificing much of its tremendous computational capacity and memory. They may choose to wander the galaxy in search of novel sources of information\cite{MGDS1981}. During their long interstellar voyages most of the systems on their ship may be suspended with little waste heat being produced. Until we attain the level of a KT-II civilization ourselves our chances of detecting these ships is virtually zero unless one shows up directly above the Earth tomorrow. We are at liberty to suspect their possible existence due to the lack of robust explanations for the missing baryonic dark matter. \section{CONCLUSIONS} \label{sect:Conclusions} Nanotechnology based self-replicating machinery may disassemble planets and turn entire solar systems into optimized computronium. The rate at which this occurs depends upon the star type, the masses of the natural objects and the element requirements and orbits of the final computing elements. Systems with large numbers of small bodies (e.g. asteroid belts), or small planets located near stars, may theoretically enshroud their stars in weeks to months. Systems with more massive or metal poor bodies will require longer periods to accomplish this. The time it takes to disassemble large planets, redistribute the matter in solar systems and breed optimal element mixes for computronium are the constraints on interstellar colonization, not the speed-of-light and high costs of interstellar travel as is more commonly thought. Advanced civilizations based on an optimized computronium infrastructure have little need for conversations with mere humans or even human civilizations whose thought capacities are trillions of times less than their own. In contrast, they may have an interest in leaving our civilization to its own unique development path so as to increase the potential diversity of, and information content in, the galaxy. This is due to the large phase space of what can be constructed using molecular nanotechnology and the difficulties in proving that the computational architectures previously adopted to support advanced civilizations are, in fact, ``optimal''. Advanced civilizations may need less developed civilizations for the ``dumb luck'' they may have in developing an unexplored quadrant of the phase space of what may be designed and assembled in support of the evolution of intelligence. The billions of large telescopes advanced civilizations may construct allow them to observe the observable regions of their galaxy at very low cost. They may also identify and communicate with civilizations that have information regarding the locations of objects that are invisible from their location. Such information is of great value in calculations of the long term motion of objects in the galaxy and is essential for civilizations seeking the lowest cost sources of additional matter and energy as well as seeking to avoid galactic hazards such as black holes or supernovas. This is due to the large energy cost and long time periods required to alter the course of entities with the mass of solar systems. The accelerating pace of technology development seems to be driving us towards the singularity\cite{VV1993}. An ultimate manifestation of this will be the conversion of our solar system into a Matrioshka Brain. Even if \emph{Rare Earth}\cite{PDW2000} is correct and we are one of a few rare intelligent technological civilizations in the galaxy, implying that both CETI and SETI will fail, and even if the missing baryonic dark matter and gravitational microlensing observations have perfectly ``natural'' explanations -- the development path outlined here merits further study as our civilization seems following it. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \acknowledgments %>>>> equivalent to \section*{ACKNOWLEDGMENTS} I would like to acknowledge Anders Sandberg for his groundbreaking explorations of megascale intelligent superentitites and thank the many Extropians and transhumanists who have made fruitful contributions to these ideas. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%% References %%%%% \bibliography{report} %>>>> bibliography data in report.bib \bibliographystyle{spiebib} %>>>> makes bibtex use spiebib.bst \end{document} %%%%