Willdemaine.co.uk

If there is no silicon heaven then where do all The debate on the existence of an afterlife for humans has been raging forcenturies, but there has been little discussion on the existence of an afterlifefor robots or electronic devices. This paper aims to examine some of thetheories on the possible existence of silicon heaven from a philosophical andscientific perspective rather than a theological one. The paper finds that it isnot possible to conclusively prove or disprove the existence of a silicon heavenwithout first answering some long standing questions on electronic life andmorality.
The sci-fi comedy Red Dwarf raises an interesting philosophical question; isthere an afterlife for machines? In the episode ’The Last Day’, the androidcharacter Kryten first describes silicon heaven: ”It’s the electronic afterlife! It’s the gathering place for the soulsof all electonic equipment. Robots, calculators, toasters, hairdry-ers – it’s our final resting place.” The question of whether there is a silicon heaven is essentially composed of two separate philosophical questions. Firstly; can robots die? Secondly;is there a heaven for them? This paper will examine both questions to determine if a silicon heaven can exist.
To prove or disprove the existence of silicon heaven we must first under- stand what it is. The Oxford English Dictionary defines heaven as: ”A place regarded in various religions as the abode of God (orthe gods) and the angels, and of the good after death, often tra-ditionally depicted as being above the sky.” This definition of heaven gives rise to many other questions. Thus, were we to use it to answer the question of a silicon heaven, we must first answerthe question of God’s existence. In the interest of parsimony, we will avoid thequestion of a ’silicon god’ by redefining heaven as ’a perfect place to whichall good silicon based electronics go after they die’. While this definitionremoves any religious connotations from the idea of heaven, it maintains thesocial preconceptions of a heavenly place; one which is perfect and is onlyaccessible after death. This definition of heaven still relies on the concepts ofperfection, morality and life - each of which will be covered in the followingsections. The redefinition of heaven also ensures that this paper is concernedonly with the existence of a heaven for electronics (robots, androids, toasters,computers etc) and not for biological organisms. Whilst not all of theseelectronic devices may consist of silicon, the term ’silicon heaven’ has beensomewhat popularised by the aforementioned television series and will beused in this paper to refer to the concept of heaven for electronics. Section3 will discuss and evaluate a number of theories on the existence of siliconheaven.
Silicon heaven cannot exist because there is not enough energy topower it If it is possible that electronic devices can live and die then suppose the ex-periment in which the simplest possible electronic device is created and thenkilled. This device has no mobility, emissions or means of communication andis powered only by a small battery. If we assume that when the battery canno longer produce an electrical output the device has died, then when thishappens there is no physical change in the constitution of the componentsof the device. The device weighs exactly the same before and after deathand exists in the same location - its physical properties remain entirely un-changed. Leibniz’s identity of indiscernibles states that any two things whichare entirely undistinguishable are in fact identical [5]. Under this identity ofindiscernibles, the device is identical before and after death. There is neither a physical change nor has the robot has moved anywhere despite it being’dead’, so it cannot have gone to heaven. Were someone to suggest that, infact, it has gone to heaven, then they would be wrong because heaven is onlyaccessible in death, yet the device was here when alive, so this cannot besilicon heaven.
Even if this simple robot was considered alive and we were unable to con- clusively prove that it had not gone to silicon heaven, then let us considerthis scenario. If silicon heaven is an afterlife for electronics then we canassume that the electronics that exist there will be alive in some capacity(i.e. they can perform the same functions as they could when truly alive)and therefore must have a source of energy. Consider ’HappyBot’ - a robotwhich is capable of thinking and feeling only positive emotions. The morethat HappyBot thinks, the happier it becomes. In order to think, HappyBotneeds electrical energy and the more energy it is given, the faster it can thinkand as a consequence, the faster it becomes happier. If silicon heaven is aperfect place which robots go to when they die, then when HappyBot dies itwill be given energy in silicon heaven to make it happy. As the only limit onits happiness is energy and silicon heaven is the perfect environment, thenit would need to be given an infinite amount of energy to make it infinitelyhappy. An infinite amount of energy is not possible based on the laws of theconservation of energy, therefore a perfect silicon heaven cannot exist.
However, both the previous arguments are based on the premise that for a robot to be alive it must have a source of electricity. Surely electricity doesn’tconstitute life else what happens when you put electricity through somethingwhich we already know to be alive? Does it become more alive? Intuitivelythis seems preposterous so electricity surely doesn’t constitute life.
Conversely, all biologically living things require some form of energy to perform the functions that define them as life. Without an energy source,organisms would be unable to reproduce, adapt or metabolise - some of theproperties which McKay (2004) deems to constitute life. If we were to createa robot which had all of these properties and could only perform them withan electrical power source, would the presence of electricity not differentiatebetween the machine being alive and dead? Silicon heaven cannot exist because robots are not alive.
The concept of death presupposes that there is life. If robots cannot livethen question of silicon heaven becomes simple; robots cannot die and as a result can not go to silicon heaven. We must therefore defend the argumentthat robots can live in order to show that it is possible that silicon heavencan exist.
The question of robot life is not a simple one as life - even biologically - is not readily definable in unequivocal terms [4]. The Oxford English Dictionarydefines life as: ”the condition that distinguishes animals and plants from inor-ganic matter, including the capacity for growth, reproduction,functional activity, and continual change preceding death.” [9] This definition of life seems intuitive as we are all familiar with the process of human life. Interestingly, the definition refers to life as a condition, whileother sources such as McKay (2004) refer to life as a process. From bothterms, it is clear that life is not something physical which is possessed, rathera characteristic which is described by a number of phenomena.
It is important to distinguish life from the idea of intelligence and con- sciousness. The one-off Red Dwarf character ’Talkie Toaster’ serves as anexemplar electronic device which shows the common blurring between intel-ligence and life. In the show, the toaster is depicted as sentient, conveysemotion and even expresses self-awareness - at one point stating in parodyof Descartes: ”I toast, therefore I am” [8]. Although the device possessesall of these qualities, it does not reproduce or grow and cannot therefore berendered as alive under the OED definition of life. However, the charactersin the show (including the toaster itself) consistently refer to the toaster withthe pronoun ’he’ which is often reserved for living organisms. Furthermore,it reflects on its previous destruction as ”first degree toaster-cide”, implyingthat it was killed (and presumably then, that it believed it was or is alive)[8].
Just as consciousness does not imply life, life clearly does not imply con- sciousness. Even a sentient robot with intelligence surpassing that of a humancannot be considered alive if it does not exhibit the phenomena which classifylife. If this robot is not alive then it cannot die and therefore cannot go tosilicon heaven. To disprove silicon heaven we must still consider all potentialforms of life, not just intelligent or conscious ones.
The computer virus (from here on just ’virus’, a biological virus will be explicitly identified) is an interesting concept which some renowned scien- tists such a Stephen Hawking have concluded that some entities described asviruses should be classified as alive: ”I think computer viruses should count as life. I think it sayssomething about human nature that the only form of life we havecreated so far is purely destructive. We’ve created life in our ownimage.”[6] It can be argued that viruses exhibit all the phenomena described in the OED definition of life. Firstly and most evidently, viruses must be ableto reproduce by definition. Many viruses are able to create exact copies ofthemselves and spread to other machines. Furthermore, metamorphic virusesare able to mutate their code upon reproducing to remain difficult to detector to spread to new systems - a process which is analogous to the mutationof biological viruses [2, 10]. Secondly, viruses often show signs of functionalactivity, whether that is actively disrupting the use of software or remainingundetected and gathering information. It can be argued that growth - inabstract terms - can be attributed to the virus during its mutation as thisleads to the virus becoming more mature and more complex. Finally, thismutation and growth implies that viruses show ’continual change precedingdeath’. If all of these premises hold then it can be said that viruses are aliveand as a consequence, that they can die and possibly go to silicon heaven.
Some may argue that a virus being alive does not imply that it can die and that deletion or modification of the viruses’ compiled code does notconstitute killing the virus. Let us consider the experiment in which thereexists both a small multi-cellular organism which is considered alive and avirus which is compiled to an executable form. Both of these entities havedefinitive constituent parts; the organism has cells and the virus has bytes.
If we were to replace a single cell of the organism with an inanimate materialof the same size and shape, the organism would continue to live, much in thesame way that replacing a single byte of the virus with a ’null byte’ - onewhich is universally accepted to perform no action - may not stop the virusfrom functioning (and thereby exhibiting signs of life). However, if we wereto iteratively replace cells of the organism with the inanimate material at afaster rate than the cells were being created, the organism would eventuallybe unable to function and would then be considered dead. In the extremecase, we could even say that the organism has only died once every singlecell has been replaced.
If the same process is repeated for the virus, it would eventually stop exhibiting its signs of life and would therefore also beclassified as dead.
Critics argue that biological viruses are only ’organisms at the edge of life’ rather than being alive in their own right due to that fact that they cannotreproduce without the use of a host [11]. While the same argument holds forcomputer viruses, it only does so by embellishing the requirements for life. Abiological cell is universally considered to be alive but can also only replicategiven the right conditions; if it is not supplied with the nutrients neededfor the replication of DNA, cell division (and therefore reproduction) cannotoccur. It is clear that a cell placed in a neutral solution does not instantlybecome dead, even though it can no longer replicate. If it is possible thatviruses can both live and die, then we cannot definitively say that siliconheaven does not exist without further reasoning.
Silicon heaven cannot exist because robots have no morality All of the prior arguments in this paper have presupposed that any robotwhich has died was morally accountable and deemed good enough to be sentto silicon heaven. If silicon heaven is the place that all the good electronicsgo, then surely there exists a silicon hell for bad or immoral electronics.
This concept is often allured to in popular culture; the Futurama televisionseries occasionally features the character of The Robot Devil (also known asBeelzebot); an electronic anti-christ who punishes the immoral: ”We know all your sins . And for each one we’ve prepared anagonizing and ironic punishment!” Dennet (1997) argues that ”Higher-order intentionality is a necessary pre- condition for moral responsibility”. At the risk of becoming kitchen appli-ance heavy, let us now consider the household toaster as our electronic agent.
This toaster may well contain a silicon chip which is used to regulate tem-perature or provide some futuristic function like browning a description ofthe current weather into the bread, but is it accountable for burnt toast?The toaster simply runs its toasting program for a time period specified byan outside agent turning a dial. It is clear that the toaster has no higher-order responsibility, it has no self awareness or ability to reason. Therefore,the responsibility of the burnt toast must reside with the outside agent. Ifthe toaster can be held accountable for failing in its normal operation, thensurely a toaster which does not burn toast can be seen as moral and cantherefore go to silicon heaven. However, just as Dennet (1997) argues thatDeep Blue should ultimately receive the credit for it’s beating of Kasparovin their first chess encounter rather than the people it who programmed it,should a toaster - aware of its heat output and the consistency of bread - not be to blame for burning my crumpets rather than the person who made orused it? If the toaster has no moral responsibility then the question of it being good or bad - moral or immoral - becomes moot. As silicon heaven is theplace where all good electronics must go after they die, we can only disproveits existence by showing that at least one good electronic device cannot gothere. If the toaster in this scenario is neither good nor bad then it doesnot qualify for silicon heaven and we must consider a robot that has moralresponsibility.
The idea of defining moral responsibility is difficult in humans let alone in robots. In the previous example it can be argued that a toaster burningtoast may constitute good behaviour if the person who programmed it wantedburnt toast. This is analogous to the more consequential scenario in which arobot is programmed to kill. If it performs its duty is it a bad robot for killingor a good robot for performing its duty? Bringsjord (2009) argues that itwould be immoral for a robot not to kill if in doing so it would be guaranteedto stop the ”incineration of a metropolis” - presumably one which is occupiedby human life [1]. Whilst this view is certainly philanthropic, it falls downon its underlying basis in pre-determinism. How would the robot ensure thatallowing the destruction of a city would not save more lives further down thecausal chain? This paper first examines the possibility of a silicon heaven under the premisethat electricity is a catalyst for electronic life. It shows that if that is the caseand that it is possible for robots to both live and die, then it is not possible fora perfect silicon heaven to exist because of the energy constraints of roboticlife. Furthermore, if the identity of indiscernibles holds, then it is shownthat silicon heaven cannot exist for physical electronic life as the physicalproperties of robots before and after death are identical.
The paper then examines whether silicon heaven can exist regardless of whether electricity is necessary for robotic life. It shows that if robots canbe said to be alive then they can be shown to die by iteratively replacing allof their constituent parts with inoperable parts until they cease to performthe phenomena which classifies them as alive. Because there are conceptualnon-physical robots (computer viruses) which can be seen as alive and have the ability to die, it is not possible to disprove the existence of silicon heavenas it cannot be conclusively shown that these robots do not go there.
The paper finally shows that if robots can both live and die, then they must be moral to reach silicon heaven. If robot morality is not possible, then siliconheaven cannot exist by definition, as it is the place where all good robotsmust go after death. If robot morality is possible, then we cannot definitivelydisprove the existence of silicon heaven until it is proven that a good robot hasnot gone there. The advancement of computer science, artificial intelligenceand biological computing is likely to generate large amount of debate on thetopics of electronic life, morality and the afterlife which this paper can onlytouch on.
[1] Bringsjord 2009, ”Unethical but Rule-Bound Robots Would Kill Us All”, [2] Cohen 1987, ”Computer Viruses: Theory and Experiments”, Computers [3] Dennet 1997, ”When HAL Kills, Who’s to Blame? Computer Ethics”, in HAL’s Legacy: 2001’s Computer as Dream and Reality, D. G. Stork (ed),Cambridge, MA: MIT Press.
[4] Emmeche 1997, Defining Life, Explaining Emergence.
[5] Forrest 2011, ”The Identity of Indiscernibles”, The Stanford Encyclopedia of Philosophy (Spring 2011 Edition).
[7] McKay 2004, ”What Is Lifeand How Do We Search for It in Other Worlds?”. PLoS Biol. 2 (2(9)): 302.
http://oxforddictionaries.com/definition/life (accessed 18/04/2012).
[10] Pan & Li L & Du & Li M & Cao & Sheng 2007, ”Differences of YMDD mutational patterns, precore/core promoter mutations, serum HBV DNAlevels in lamivudine-resistant hepatitis B genotypes B and C”.
[11] Rybicki 1990, The classification of organisms at the edge of life, or prob- lems with virus systematics. S Afr J Sci.

Source: http://www.willdemaine.co.uk/assets/files/essay.pdf