Up to the Transhuman PageAs Dyson pointed out in "Physics and Biology in an Open Universe" , most physicists are very daring in modelling the past, but very few dare to model the future evolution of the universe. In that paper he did just that, and also realised that life may become an important part in the equation. Tipler has developed his own scenario in more detail, and even if his theories fail I think we have good reason to regard Dyson and Tipler as among the founders of the field of "physical eschatology", the study of possible future evolution of the universe and how we might affect them.
Tipler have come under a lot of attack for his theories, not entirely undeserved. Applying theology to physics (or the reverse) have long been a taboo, and he boldly breaks it. Whether this is a good or bad thing is hard to tell; there is enormous potential for misunderstanding, confusion and pure pseudoscience, but at the same time it might force some fresh air into the stale dialogue between faith and science. As I see it, Tipler has made several quite audacious claims which probably don't hold water, but many of his ideas and lines of reasoning are probably sound, and might one day be used for other purposes than to prove the existence of the personal God Tipler seeks.
In the following I will give an overview of his theory and arguments, and my own comments to them. I will try to avoid overly technical details or place them on their own footnote-pages. I will also not go into the discussion about identity, emulation and the many-worlds interpretation of quantum mechanics which fills much of PoI, it has been (and is) discussed elsewhere and I agree more or less with Tipler on many of these points.
II The amount of information processed along the curve between now and the c-boundary is infinite. (Life evolves forever)
III The amount of information accessible to beings on gamma stored around it diverges as it approaches the c-boundary. (Life never gets trapped in the Eternal Return and remembers its past).
In Tipler's theory the universe is closed and will collapse back into a final pointlike singularity, so the c-boundary will be a point with the above properties. In order for this to happen, the universe has to collapse in a very specific way, not too fast (in that case life in different parts of the universe cannot exchange signals an infinite number of times in order to obey III), not too isotropically (since life will need the shear energy of space as an energy source) and not too anisotropically (since if horizons form the amount of information accessible to a point is limited). In addition, as the universe grows hotter and hotter, life has to evolve to work at ever higher energy levels, which means that there have to be suitable particle spectra to use as a basis for existence and to store the ever accumulating information.
This is his classical description of the Omega Point. Tipler have shown that the laws of physics apparently do not rule out this kind of future evolution of life under certain circumstances, which is interesting in itself. He then goes further, and claims that the Omega Point will actually happen, as the result of natural law. This is the Omega Point Theory.
The Eternal Life Postulate claims that life will exist forever. If one accept it, Tipler claims one have to accept the Omega Point since it provides the only possible way to bring this about (I would rather say it provides the only known or conjectured way to do this, there might be others). Since life in general seems to evolve towards increased changes of survival in any given environment, one could argue that evolution will favour this Postulate, even if it is not a ontological truth. However, we should always remember that evolution is blind, and many species or groups of organisms have become trapped in blind alleys (like the insects, whose size cannot increase due to their tracheatic system). It is possible that life (and intelligent thought) may become trapped in such a way as to make eternal life impossible. However, the postulate is just one of the reasons Tipler accepts Omega as inevitable.
The Quantum Omega Point Theory of Tipler is "Quantum Gravity with the Omega Point Boundary Condition". Tipler accepts the Many Worlds Interpretation of Quantum Mechanics, and claims that all existing universes are parts of an universal wave function extending through superspace. The Omega Point forms its boundary condition; any universe which do not end in It simply does not exist. Since his Omega Point requires some (actually a lot) fine tuning, it would be falsifiable. In PoI he shows that it requires the flowing conditions to hold:
But it is quite easy to imagine many possible Omegas: one universe where Omega is a single, benevolent entity resurrecting humans and allowing them to live in paradise forever, another where Omega is a complex program whose sole purpose is to survive, and a third universe where different entities compete and evolve forever, without ever reaching stability.
In fact, Tipler implicitly assumes that the universe will reach a limiting state. But this might never happen, since there is infinite subjective time for change and growth if life can develop indefinitely. The boundary condition could be a certain behaviour, not a state.
Mitch Porter writes:
Cosma Shalizi writes, in the Aleph anti-FAQ, that "in a universe going on forever, there is infinite time for error, coincidence and malice." By the same token there is infinite time for forethought and error avoidance.
I find this interesting, but not convincing. After all, one could equally well have the universe divided into two, three or a billion separate god-beings who evolve more and more, without ever merging into a single whole. And as the resources available to life grow in the final phases, it might instead begin to diverge in an infinite number of directions. In addition, since there is infinite subjective time to change in, the omega-beings may split and merge an infinite number of times.
It is also a bit doubtful from a theological standpoint, since most (western) theologians claim that God exists outside the universe and physics. But Tipler correctly points out that the universe is by definition all that exists, so either God is the universe or a part of it, or He doesn't exist. However, the space-time continuum we exist in (and usually call the universe) doesn't necessarily include God, who could be a part of some other aspect of the universe.
Tipler's identification of the Holy Ghost with the Universal Wave- Function made me laugh aloud, but I think he has a point here. It fits the concept of "God as pure existence" very well. It is however not the personal God Tipler seeks.
Tipler attempts to solve the theodicy problem by invoking the
MWI: since all possible histories exist, all logically possible evils
must also happen. There might be histories with less evil or none at
all, but they have a very small measure. And these evils are also
logically necessary, since all histories have to lead to the Omega
Point in the end if the OPT holds. Tipler's Omega-God cannot do
anything about them, since they are necessary and the universe
would not exist without them. Essentially Tipler's solution is to
avoid the problem of most theodicies where God just have created
one world, and could have made it different.
Resurrection
Tipler claims that Omega will resurrect everyone into an immortal
life in what could only be called paradise. That a future
superintelligence could resurrect all dead beings from earlier times
using scientific means (Universal Immortalism) has already been
suggested by Hans Moravec in "Mind Children", so the idea isn't
new. In Tipler's scenario, the immense computing resources of the
final stages of the Big Crunch would make it possible to not only
simulate the previous history of the universe, but also to
emulate it, which would allow the resurrection of all beings who
have ever lived (and probably all beings who could have
lived).
This form of resurrection requires that the Strong AI Postulate and the "No-Clone Theorem" holds. The issue of preservation of identity through uploading and similar processes also comes up. These questions are already hotly debated in other areas.
The big problem with universal resurrection is whether it will really happen; why would Omega bother to resurrect myriads of lesser beings? Tipler argues this from his assumption that Omega will be extremely altruistic and give us a chance to develop indefinitely. He argues this from simulations of the Prisoners Dilemma , which show that altruistic strategies can out-compete selfish strategies by cooperating with each other; in the far future of the Omega Point this would have led to a situation where everything is completely altruistic.
Unfortunately this rests on a few shaky assumptions. Other simulations of the Prisoners Dilemma has shown successive stages of altruism and selfishness; a population of selfish systems is less effective than groups of altruists which help each other, so in time altruists appear and out- compete the selfish strategies (who spend most of their time defecting against each other). Then a situation where most strategies are altruists develop, but now parasites and defector- strategies have plenty of suckers to fool, and the situations devolves back into the defector stage, and the cycle can begin anew. This would imply that in the Omega aeon there could be an infinite number of "wars in Heaven" between altruists and selfish entities, never resolving into the static altruist paradise of Tipler.
Tipler also shows that for beings near the Omega Point, resurrecting all beings of the previous universe is actually not very demanding (this is roughly the same economic argument he uses to argue that von Neumann probes will eventually become cheap enough for private launch), so even if there are many different beings with different goals, some altruist might do it. While this is nice, and somewhat more believable, it is not very reassuring since selfish or sadistic beings also could resurrect us all into eternal hell. Tipler seems to be totally convinced that good will always triumph in the end and that there will be just one cosmic being, but this is a very shaky assumption. There might be alternatives to good and evil we do not yet know about.
Tipler doesn't seem to acknowledge mainstream immortalism, which tries to make us immortal in the near future instead of relying on the altruism of a future super-intelligence. While his idea of a future resurrection might be comforting, at the same time it does not encourage individual seeking of self-transcendence, since it will be supplied by the Omega Point in the future.
Tipler then argues that intelligent life might make use of quantum randomness in their decision-making processes (the q-mind theory goes further, and claims quantum effects have a crucial role), and thus get free will. Tipler argues this from the assumption that mixed strategies have an evolutionary advantage for living beings. Mitchell Porter has pointed out that even if this is true, natural selection still have to come up with the solution; diamond is the hardest substance in nature, but we don't have diamond skeletons. It is worth noting that Tipler's chain of reasoning also acknowledges free will to inanimate objects which are influenced by quantum processes such as the box in the Schrodinger's cat experiment.
It is also worth noting that there are some opponents to Strong AI (especially Penrose) who argues that the human mind is able to make true noncomputable deductions; if this is true then such minds could at least in principle circumvent the non-classification theorem and determinism would hold!
However, even if life doesn't use quantum noise, Tipler claims there is free will in the QOPT since the free actions of intelligent beings will lead to the OP which in turn defines the laws of physics (through the Omega Point Boundary condition and the self- consistency of the laws of physics). I cannot make heads or tails of this argument, it seems to be totally circular logic and contradicts the inevitability of the universe he invokes in his solution of the theodicy problem. If there is no free will, then agents will make their deterministic choices which will lead to the OP, and this would be consistent too.
Unfortunately, as I have outlined above, it is based on many assumptions and might even be seen as containing a lot of unnecessary additions (why has the boundary to involve superintelligences? Could the c-boundary not consist of "natural" anisotropic collapses?). One could accuse Tipler of having devised a model with very desirable properties, and then formed a theory around it, instead of the other way around. In fact, Tipler repeatedly invokes the Omega Point Boundary Condition to remove threats to his theory; the possibility that quantum fluctuations could wipe out all life near the singularity or that some horizons are unremovable are simply dismissed since they would contradict the theory!
Tipler points out that infinite progress is impossible to achieve in eternal chaotic inflation cosmologies, where life could escape the heat-death of their old domains by forcing the inflation of new domains or finding the entrances into them, since the amount of information which can be transmitted into them is finite due to the Bekenstein Bound (which is not proven for strongly curved spacetimes but might hold in a modified form). He also points out that this holds for creating "basement universes".
However, this might be a viable method of escaping an open universe into an universe where an Omega Point might take place (although this possibility is outside the OPT, since it assumes the Omega Point Boundary Condition on all space-time). Even if one cannot move an arbitrary amount of information into the new domain, this will not make the creation of an Omega Point impossible (or even rule out a resurrection; Tipler has pointed out that a brute-force resurrection could be done even if nothing is known about the previous states by emulating all possible states).
John Walker has pointed out that black holes pose a serious problem for the Omega Point; their event horizons would prevent the infinite growth of information required for the OP, both by denying access to part of space-time and because of information loss into the singularity. Exactly what happens to information lost into black holes is hotly debated, and may or may not be a serious problem. But the event horizons certainly pose a problem for the OPT. Somehow all black holes have to be removed from the universe before the singularity. This could happen through Hawking decay, but that would rule out the rather small closed universe Tipler favours.
Another big hole which Tipler handwaves away is how computational processes can be done at the increasing energies near Omega. While some critics have dismissed the idea of such processes out of hand, I think that is premature. We know that Turing-equivalent computers can be built out of model trains, game of life patterns, ropes and pulleys, silicon or bouncing billiard balls, so why not plasma too? The difficulties seem to be more a matter of vision and understanding than physical impossibility; we simply have too little experience to envision a computational system in a regime where quantum effects and fields play a major role. In order to gain credibility the OPT has to include at least an existence proof that such a system is possible.
A similar problem is that Tipler does not demonstrate that intelligent life could force a Taub-like collapse, which is needed for this scenario. He refers to the chaos in the relativistic field equations, but never mentions an actual way to manipulate the evolution of the entire universe on this scale. Presumably this could be done by moving large quantities of matter and energy purposefully, but this would probably require long-range cooperation despite the fact that the cosmic antipodes are unable to communicate with each other until either the collapse have been manipulated or the universe collapses into the big crunch. Another problem with this is that actually moving matter on these scales might take too long time for his model (in ACP he estimates the time needed for intelligent life to artificially disrupt a galaxy to 10^22 years, which is clearly outside the bounds of the scenario he describes in PoI).
He also promises several proofs which have so far not been given, especially of the Eternal Life Postulate, although Tipler might claim his theory is a proof of it (leading to a circular loop).
One intriguing thing about the QOPT is the role of Big Bang in it. Tipler claims the universe is not contingent, and that the OP forms a boundary for it. Through backwards causality it (the "Alpha Point") is necessary to bring the universe into being as seen through our timescale. But maybe we could view it as an Omega Point seen backwards? It would be a part of the boundary of the universe, and intelligences moving backwards in time (based on backwards causality) would evolve towards infinity as it is approached. In fact, if the QOPT and the "No Clone Theorem" holds, they would be identical. This is both symmetric, elegant and rather bizarre.
In many places Tipler mixes up his arguments in PoI, accidentally (?) confusing simulations with emulations (e.g. p 210), the Omega Point with the beings in the later eras of the universe and his theory with observable reality.
Another thing which irritates me a bit is that Tipler takes a lot of things for granted about all possible forms of intelligence (even if there are only humans in the universe, by the time we have developed to the level of technology that this theory becomes relevant we will certainly have diverged greatly from the present). For example, he assumes immortality and progress are the highest good to all beings, which might not be the case (there is an evolutionary advantage in diversity and survival of ones genotype, but the rest is a matter of taste). He also attacks the Eternal Return with "ad hominem arguments" by linking it with Nazism - would that really be relevant to an alien being?
While the theory does not really hold water, I think it is worth attention as a first attempt. Maybe we one day can develop a similar but stringent theory, or use this vision to make the Omega Point an engineering possibility (the transhumanist perspective). Science, Philosophy and Theology might not be completely incompatible, but I think we still are a long way off from "physical theology" (and an even longer way from a widespread acceptance of it).
I also thank the other members of the Omega Point Theory Mailing List for stimulating discussions which have inspired me to large parts of this document.
Up to the Omega Point PageAnders Sandberg / asa@nada.kth.se