The structure of the global catastrophe

Indirect estimation of probability of natural catastrophes

If not to consider effects of observant selection, we receive very good chances on вживание in the XXI century from any natural (but not anthropogenous) catastrophes - from galactic to geological scales as from this, that they were not during existence of the earth and our kind, very small probability of follows that they will occur in the XXI century. As any natural catastrophe has not destroyed human ancestors for the last 4 billion years, it is possible to conclude, that our chances of  destruction the XXI century from natural catastrophes make less than 1 to 40 million. (And with the account of high human survivability and приспособляемости, and it is less than that.) Unfortunately, such reasonings are essentially incorrect, as do not consider малоочевидный effect of observant selection.

Owing to this effect expected future time of existence will be less, than the past (see more in detail I wash article «Natural catastrophes and Antropic principle» and the head about observant selection in section about natural catastrophes). Nevertheless hardly the contribution of observant selection makes more than one order. However for different levels of natural catastrophes we have the different characteristic periods of time. For example, the life on the Earth exists already 4 млрд years, and, with the account told, it could exist yet less than 100 - 400 million years. (Observant selection here consists that we do not know, what share of planets of terrestrial type perishes in the course of evolution; having assumed, that the share of the escaped makes from 1 to 1000 to 1 to billion, and we receive estimations in 100-400 million years as a half-life period.) That is the indirect estimation of probability жизнеуничтожающей natural catastrophe would be 1 to 4 000 000 for hundred years. It пренебрежимо small size against other risks. As to time of existence of our kind last natural catastrophe threatening to it, was much closer in time, 74 000 years ago (volcano Тоба) and consequently we have expected time of existence only 7 000 years with the account of the greatest possible effect of observant selection. Observant selection here consists that if people were very little hardy kind which dies out with periodicity of times in some thousand years, we could not notice it as we can notice only that branch of our kind which has lived sufficient term for civilisation formation in which we can ask the given question. 7000 years would correspond with the account of a huge error of such reasonings about 1 % of extinction in the XXI century as a result of natural catastrophes or immanent to an instability kind - and it is the maximum estimation in the worst case. If not to take in calculation observant selection chances of natural catastrophe of any sort, leader to mankind extinction, on the basis of last time of existence it it is possible to calculate by means of formula Готта (applied by time of existence Homo sapiens), and they will be 1 to 1500 for 100 years, that is 0,066 % that is not enough.

At last, there are the third sort of the catastrophe which probability we can indirectly estimate on last time, namely, on time during which there is a written history, that is 5000 years. We can safely assert, that for 5000 years there was no catastrophe which would interrupt written history. Here too it is possible, but observant selection as operate not natural more strongly, but anthropogenous factors is improbable. That catastrophe which could interrupt written history 3000 years ago, for example, supervolcano eruption in средиземноморье, now could not make it any more. Therefore it is possible to tell safely, that the natural catastrophe interrupting written tradition (such as it was in the past, instead of now) has chances no more than 1 % in the XXI century, considering on formula Готта (applying it by all time of existence of written tradition). And as now the written tradition is much stronger, than in the past it is possible to reduce safely this estimation twice: to 0,5 %. And even such catastrophe which would interrupt writing in the past, will not interrupt it now and will not kill all people.

At last, the effect of observant selection can be shown and in the relation to anthropogenous catastrophes, namely, to global risk of nuclear war. (In the assumption, that general nuclear war would destroy mankind or would reject it so back, that the writing of books would become impossible.) The effect of observant selection here consists that we do not know what were chances of our civilisation to survive during the period with 1945 till 2007, that is during existence of the nuclear weapon. Perhaps, in nine of ten «the parallel worlds» it was not possible. Accordingly, as a result we can underestimate global risks. If intensity of change of number of observers would be very great, it would have "pressing" influence on date in which we would find out ourselves - that is we most likely would find out ourselves early enough. See more in detail article Bostromа and Тегмарка where exact calculations for catastrophes космологических scales are offered. If the probability of risk of extinction would make 90 % a year then I, most likely, would live not in 2007, and in 1946. That I am still live in 2007, gives a certain top limit (with the certain set reliability) on rate of extinction (for historical conditions of the XX-th century). Namely: 5 summer period of "half-decay" can be excluded approximately with probability 99,9 (as for 50 years there have passed 10 cycles for 5 years, and 2 in 10 degrees it is 1024. That is for 50 years one thousand share of planets would escape only.) arguing further in a similar way it is possible to exclude authentically enough the periods of "half-decay" of a civilisation smaller, than 50 years. However big we cannot exclude. It, certainly does not mean, that the real period of "half-decay" and is 50 years, however, if to start with a precaution principle should be assumed, that it so. Such half-life period would mean our chances to live till XXII century approximately in 25 %. (And it in the assumption, that level of threats remains invariable from the XX-th century middle.)

Conclusions: various independent methods of indirect reasonings give estimations of probability of  destruction of a civilisation in the XXI century in tens percent. It should not calm us in the sense that as if it guarantees to us tens percent of a survival. Because if to consider degree of uncertainty of such reasonings it is category events «tens percent» which as we have assumed in the beginning, means risks from 1 to 100 %.

Reasoning on Simulation

Ник Bostrom has developed the following logic theorem named a reasoning on Simulation (we already mentioned it in a context of risks of sudden switching-off of "Matrix" and essentially new discovery). Here a course of its reasonings:

Proceeding from current tendencies in microelectronics development, it seems quite probable, that sooner or later people will create a spontaneous artificial intellect. Nanotechnologyи promise limiting density of processors in billion pieces on gramme of substance (carbon) - with productivity of an order 10 флопс. nanotechnologyи coal deposits will allow to transform into the huge computer (as the basic building material for it, probably, there is a carbon) plus. It opens prospect of transformation of all Earth in «компьютрониум» - one huge computer. Capacity of such device is estimated in 10 operations in a second. (That corresponds to transformation of one million cubic kilometres of substance in компьютрониум which will cover all Earth with a layer in 2 metres.) Use of all firm substance in solar system will give an order 10 флопс. It is obvious, that such computing power could create detailed simulations of the human past. As it is supposed, that for simulation of one human it is necessary no more than 10 флопс (this number is based on quantity нейронов and синапсов in a brain, and frequency of their switching) it will give the chance to feign simultaneously 10 people, or 10 the civilisations similar to ours, with sew in the speed of development. Hardly компьютрониум will direct all resources on feigning of people but even if it will allocate for it one million efforts, it will be still an order 10 human civilisations. So, even if only one of one million real civilisations generates компьютрониум this компьютрониум generates an order 10 civilisations, that is on each real civilisation it is necessary 10 virtual. Here concrete figures, and that at quite realistic assumptions the set of the feigned civilisations on many usages is more than set of the real are important not.

From here Ник Bostrom does a conclusion that one statement from three is true at least:

1) Any civilisation is not capable to reach the technological level necessary for creation компьютрониума.

2) Or EVERYONE possible компьютрониум will be not interested absolutely not in modelling of the past.

3) Or we already are in imitation in компьютрониуме.

Thus point 2 can be excluded from consideration because there are reasons on which at least some компьютрониумам it will be interesting what were circumstances of their occurrence, but are not present such universal reason which could operate on all possible компьютрониумы, not allowing them to model the past. The reason of interest to the past can be much, I will name is a calculation of probability of the occurrence a little to estimate density of other supercivilizations in the Universe or entertainment of people or certain other beings.

Point 1 means, that or компьютрониум and simulations in it are technically impossible, or that all civilisations perish earlier, than find possibility it to create, that, however, does not mean with necessity of extinction of carriers of this civilisation, that is for our case of people, but only crash of technical progress and recoil back. However it is not visible the rational reasons yet, doing компьютрониум impossible. (For example, statements that consciousness simulation is impossible as consciousnesses де is quantum effect, does not work, as quantum computers are possible.) thus it is impossible to tell, that компьютрониум it is impossible basically as people it is peculiar to have dreams, not distinguishable from within from a reality (that is being qualitative simulation) so, by means of genetic manipulations it is possible to grow up a superbrain which has dreams continuously.

Thus, the reasoning on simulation is reduced to sharp alternative: «Or we live in the world which is doomed to be lost, or we live in computer simulation».

So, the  destruction of the world in this reasoning does not mean extinction of all people - it means only the guaranteed stop of progress before компьютрониум will be created. Гарантированность it means not only, that it will occur on the Earth, but also on all other possible planets. That is it means, that there is certain very universal law, which interferes suppressing (on many usages) to the majority of civilisations to create компьютрониум. Probably, it occurs simply because компьютрониум is impossible, or because modelling of human consciousness on it is impossible. But can be, that it occurs because any civilisation cannot reach level компьютрониума as faces certain unsoluble contradictions, and is compelled or to be lost, or will be rolled away back. These contradictions should have universal character, instead of to be connected only, say, with the nuclear weapon because then civilisations on those planets in which bark there is no uranium, can steadily develop. The theory of chaos which does systems above certain level of complexity essentially astable can be an example of such universal contradiction.

The known objection leans against these reasonings that reality simulation not so necessarily is a copy of that was in the past. (Whether the review of objections to a reasoning on simulation in Daniel Medvedev's article «see we Live in gamble Ника Bostromа?» ) And if we are in the designed world it does not allow us to do conclusions about what real world. As from a computer game the monster, for example, cannot guess a real peace arrangement of people. However that we do not know, what world outside of simulation, does not prevent to know to us, that all of us are in simulation. Here it is important to distinguish two senses of a word "simulation" - as computer model and as that fact, that this model reminds a certain historical event from the past. Thus it is possible to assume, that the majority of simulations are not exact similarity of the past, and the considerable share of simulations does not concern at all the past of that civilisation which then has created them. As well in the literature the majority of novels is not historical novels, and even historical novels not precisely coincide with the past.

If we are in simulation, we are threatened with all same risks of  destruction which can happen and in realities, plus intervention from authors of simulation who to us can throw certain «difficult problems» or investigate on us certain extreme modes, or simply take a fun at our expense as we have a good time, looking through films about falling of asteroids. At last, simulation can be simply suddenly switched off. (Simulation can have a limit ресурсоёмкости, therefore authors of simulation can simply not allow to create to us so difficult computers that we could start the simulations.)

So, if we are in simulation, it only increases the risks which have hung over us and creates essentially new - though there is a chance of sudden rescue from authors of simulation.

If we are not in simulation the chance is great, that any civilisations because of catastrophes do not reach creation level компьютрониума which we could reach by the XXI century end. And it means, the probability of certain global catastrophes which will not allow us to reach this level is great.

If we adhere Байесовой to logic, to us followed attribute equal probabilities to independent hypotheses. And then we should attribute to a hypothesis that our civilisation will not reach level компьютрониума 50 % of probability (that to actually equivalently statement that waits for it a certain crash in the XXI century). This estimation coincides in the order of size with estimations which we have received in other ways.

It turns out, that the reasoning on simulation operates in such a manner that its both alternatives worsen our chances of a survival in the XXI century, that is it net the contribution negative irrespective of the fact how we estimate chances of one of two alternatives. (My opinion consists that probability of that we are in simulation, - above, than probability of that we a real civilisation to which can be lost, and on many usages.)

It is interesting to note repeating pattern: the alternative with SETI also has negative net-effect - if they nearby we in dangers if they are not present, we too in danger as it means, that some factors prevent to develop it.

Integration of various indirect estimations

All resulted indirect estimations are executed independently from each other though yield about the identical and unfavourable results, consisting that the probability of human extinction is high. However as these reasonings concern the same reality, there is a desire to unite them in more complete picture. The reasoning on simulation Bostromа exists logically separately from a reasoning Carter-Lesli doomsday (which else it is necessary to connect with formula Готта), and accordingly there is their temptation to "marry". Such attempt is undertaken in work Иштвана Араньоси «Doomsday Simulation Argument». Them, in turn it is interesting to connect with multipeace immortality in the spirit of Хигго and with influence of effect of observant selection.

Interesting attempt such is undertaken in already mentioned article Кноба and Олума «Philosophical value космологической inflations». In a counterbalance «private Doomsday argument» in the spirit of Carter-Lesli, they put forward «Universal Doomsday argument». Namely, they show, that from this, that we find out ourselves in the early form of mankind, follows, with high probability, that the set of people which is in короткоживущих civilisations, is more than set of all people who are in all long-living civilisations on all Universe, or, in other words, the quantity of long-living civilisations is not enough. It besides means, that chances of our civilisation not to live millions years and not to occupy a galaxy - are rather great, however changes the probable reasons of extinction: namely, it will occur not because of any private reason, concerning only to the earth, and because of a certain universal reason which would operate on all planetary civilisations more likely. We should be anxious, they write, not an orbit of a concrete asteroid, and that in all planetary systems there are so many asteroids that it does a survival of civilisations improbable; we should be anxious not by that a certain concrete nearest star becomes supernew, and that lethality supernew is essentially underestimated. We will notice, that the same conclusion that the set короткоживущих civilisations considerably surpasses set long-living, follows and from a reasoning on simulation Bostromа if in quality короткоживущих civilisations to consider simulations.

I believe, that the essence of this integration should be that we will find out, what reasonings block others that is what of them are stronger in logic sense. (It is thus possible, that the subsequent researches can give more exact picture of integration, and will reduce all separate calculations to one formula.) I see such order of capacity of statements (stronger statements cancelling weaker, from above). Thus I do not mean, that all of them trues.

a. The qualitative theory of consciousness based on concept about квалиа. Квалиа is the philosophical term designating the qualitative party in any perception, for example, «красность». The nature and a reality квалиа are object of intensive discussions. Theories about квалиа do not exist yet, there are only a few logic paradoxes, with it connected. However, apparently, the theory about квалиа can exclude representations about plurality of the worlds and linearity of time. Owing to it such theory, it be created and proved, would make unauthorized any below-mentioned reasonings.

b. A reasoning on immortality of J. Хигго, based on idea about plurality of the worlds. In this case always there will be a world where I, and the part of a terrestrial civilisation accordingly, was not lost. The reasoning on immortality Хигго is very strong because it does not depend neither on a doomsday, nor from, whether there are we in simulation or not. Immortality on Хигго does a humanal doomsday impossible. Any owner of simulation cannot affect work of reasoning Хигго in any way because always there will be an infinite quantity of other simulations and the real worlds, in accuracy coinciding with given in time present situation, but having with it the different future. However reasoning Хигго leans on «self sampling assumption» - that is idea that I are one of copies of set of the copies - and all subsequent reasonings lean against the same idea - argument about simulation, formula Готта, a reasoning on Carter-Lesli doomsday also. Any attempts to deny immortality on Хигго, based on impossibility of consideration of as one of copies of set of the copies simultaneously deny also all these reasonings.

c. A reasoning on simulation Bostromа. It too works in the assumption of plurality of the worlds whereas the subsequent reasonings do not consider this fact. Besides, if we actually are in simulation we do not observe the world during the casual moment of time as simulations, more likely, will be adhered to historically interesting epoch. At last, reasonings in the spirit of DA demand possible continuous numbering of people or time, that in case of set of simulations does not work. Therefore any forms DA become invalid, if the reasoning on simulation is true. The reasoning on simulation is stronger than a reasoning on Carter-Lesli doomsday and formula Готта because it works, irrespective of, how many still people will be in our real world. Moreover, it essentially washes away concepts about quantity of people and volume, that such the real world as it is not clear, whether we should consider the future people from other simulations, as real. Not clearly also, whether each simulation should feign all world from the beginning up to the end, or only a certain piece of its existence only for several people.

d. Formula Готта. Formula Готта confidently works concerning the events which have been not connected with change of number of observers, for example, concerning radioactive disintegration, date of a pulling down of the Berlin wall, a prediction of duration of a human life etc. However it gives much softer estimation of the future duration of existence of mankind, than Carter-Lesli argument. Formula Готта is more simple and clear tool for a future estimation, than Carter-Lesli reasoning. At least because formula Готта gives concrete numerical estimations, and Carter-Lesli reasoning gives only the amendment to initial probabilities. Further, formula Готта is applicable to any referential classes as for any class it gives an estimation of time of end of nominal this class. And in Carter-Lesli reasoning the death of the observer is mentioned usually, and he should be adapted to situations where the observer does not die. Question on, whether it is necessary to apply the amendments given by a reasoning of Carter-Lesli to estimations which has given formula Готта, demands the further research.

e. Carter-Lesli argument. The important condition of argument of Carter-Lesli (in its interpretation Bostromом) is несуществование other civilisations, besides terrestrial. Besides, it is very difficult to think up real experiment in which it would be possible to check up force of this reasoning. And mental experiments work with certain stretches.

f. Fermi's paradox is too in the bottom of this table as a reasoning on simulation evidently cancels its value: in simulation the density of civilisations can be any, no less than risk of their aggression, depending on whim of owners of simulation.

All told here about indirect ways of an estimation is on the verge between demonstrable and hypothetical. Therefore I suggest not to take on trust made to a conclusion, but also not to reject them. Unfortunately, researches of indirect ways of an estimation of probability of global catastrophe can throw light on our expected future, but do not give keys to its change.

Chapter 25. The Most probable scenario of global catastrophe

Now we can try to generalise results of the analysis, having presented the most probable scenario of global catastrophe. It is a question not of an objective estimation of real probabilities which we can calculate only concerning falling of asteroids, and about value judgment, that is «best guess» - the best guess. It is obvious, that such estimation will be painted by humanal preferences of the author, therefore I will not give out it for the objective precisely calculated probability. Depending on what will appear the new information, I will correct the estimation.

In this estimation I consider both probability of events, and their affinity to us on time. Therefore I attribute small probabilities nanotechnologyческой grey goo which though it is possible technically, but is eclipsed by earlier risks connected with biotechnologies. Precisely also creation of the nuclear Doomsday Machinetoo demands many years and is economically inexpedient, as the damage of such scale more cheaply and will faster put by means of the biological weapon.

These assumptions are made concerning offered threats even with the account of that people will try to resist to them so, how much it can. So, I see two most probable scenarios of global definitive catastrophe in the XXI century, the leader to full human extinction:

1) the Sudden scenario connected with unlimited growth of an artificial intellect which has unfriendly concerning human of the purpose.

2) the System scenario in which the leading part is played by the biological weapon and other products of biotechnologies, but also will be used the nuclear weapon and microrobots. Will play also the role distribution of superdrugs, pollution of environment, exhaustion of resources. The essence of this scenario that there will be no one factor destroying people, and will be a shaft of set of the factors, surpassing all possibilities on a survival.

The most probable time of action of both scenarios - 2020-2040. In other words, I believe, that if these scenarios are realised, more than 50 % they will occur to chances in the specified time interval. This estimation occurs from this, that, proceeding from current tendencies, hardly both technologies will ripen till 2020 or after 2040.

Now we will try to integrate all possible scenarios with the account of their mutual influence so that the sum was equal 100 % (thus it is necessary to consider these figures as my tentative estimation to within an order). We will estimate the general probability of human extinction in the XXI century, according to words of sir Martin Risa, in 50 %. Then following estimations of probability of extinction seem convincing:

15 % - unfriendly AI or struggle of different AI destroys people. I attribute AI such high probability because AI possesses ability to find and influence all people without an exception - in большей to a measure, than other factors.

15 % - system crisis with repeated application of the biological and nuclear weapon.

14 % - something unknown.

1 % - uncontrollable global warming and other variants of the natural catastrophes caused by activity of human.

0,1 % - natural natural catastrophes,

0,9 % - unsuccessful physical experiments

1 % - grey goo - nanotechnologyческая catastrophe

1 % - attack through SETI

1 % - the nuclear weapon of the Doomsday

1 % - other.

The remained 50 % fall to chances of that in the XXI century people will not die out. They see consisting of:

15 % - positive technological Singularity. Transition to a new stage of evolutionary development.

10 % - Negative singularityin which course people survive, but lose value. Variants: survived in the bunker, a zoo, the unemployed at the TV. The power passes to AI and cars.

5 % - Sustainable development - the human civilisation develops without jumps in technologies and without catastrophes. It is offered as the best variant traditional futurologists.

20 % - Recoil on a stage постапоклиптического the world. Different levels of degradation.

Now we will consider possible influence on these figures of different forms of the theorem of a doomsday. Formula Готта taken concerning all quantity of people on the Earth, gives not so an extinction every prospect in the XXI century - at level of ten percent, however considerably limits chances of mankind to live a next millenium or больший term.

One more variant of reasonings with use DA and formulas Готта consists in its reflective application - and legitimacy of such application is seriously challenged. Namely, if to apply formula Готта to my rank (that is number by date of occurrence) in set of all people which know about formula Готта or DA it will be soon definitively denied, or chances of a survival in XXI century appear illusive. It is connected by that one of the most extreme and disputable decisions of a problem of referential classes whom concerns DA, consists that it concerns only those people who know about it - and such decision of a problem of referential classes was offered still by pioneer DA B.Karter when for the first time reported about DA at session of the Royal society. Extremeness of this decision that as in the past is a little people who know DA (about ten thousand at the moment), that fact that I find out myself so early in this set, speaks, agrees to the logician most DA, as in the future will be approximately people as much knowing about it. As the number knowing about DA is continuous нелинейно grows, already through some tens years it should reach millions. However, it agree to the logician most DA, it is improbable, time I have found out myself so early in this set. Hence, something will prevent that the set knowing about DA will reach such big size. It can be or refutation DA, or that will not be simple people who will be interested in it. As well as many other things can be denied variants DA, this variant, having specified that I am not casual observer DA during the casual moment of time, and certain features a priori inherent in me have led to that I am interested in different unchecked hypotheses at early stages of discussion.

Carter-Lesli reasoning does not give a direct estimation of probability but only modifies an aprioristic estimation. However the contribution of this updating can be so considerable, that the concrete size of an aprioristic estimation of probability is not becomes important. For example, J. Leslie results the following example of application of a reasoning of Carter-Lesli in the book: aprioristic probability of extinction in the near future in 1 %, and rupture between number of mankind at "bad" and at the "good" scenario in one thousand times. Then these aprioristic 1 % turn at it through formula Байеса in апостериорные 50 %. However if we apply the same assumptions to our aprioristic probability in 50 % we will receive chances of extinction in 99,9 %.

At last, the third variant of the Theorem of the Doomsday in formulation Bostroma-Tegmarka adapted by me to less scale natural processes, does not render essential influence on probability of natural catastrophes in the XXI century as limits degree of underestimation of their frequency to one order, that all the same gives chance of less than 0,1 %. The worst (but) display of effect of observant selection underestimation of probability of global nuclear war which would lower the maximum frequency of this event from one event of times in some tens years, to one event of times in some years would be absolutely not obligatory. Nevertheless the top border is yet value so here all not so is bad.

So, indirect ways of an estimation of probability of global catastrophe or confirm an estimation of an order of 50 % in the XXI century, or sharply increase it to 99 % ­ - however those variants of reasonings in which it sharply increases, do not possess as much high - 99 % - validity degree. Therefore we can stop on a total estimation in «more, than 50 %».

Much easier to think out scenarios of global catastrophe, than ways of its prevention. It suggests that the probability of global catastrophe is rather great. Thus all described scenarios can be realised in xxi a century. Ник Bostrom estimates probability of global catastrophe as «not less than 25 percent». Martin Rees - in 30 percent (for 500 next years). In my subjective opinion, it more than 50 percent. Thus it погодовая the probability is more than 1 percent also grows. The peak of this growth is necessary on first half of XXI century. Hence, very many depends on us now.

At the same time to predict the concrete scenario at the moment it is unreal, as it depends on set of unknown humans and random factors. The number of publications on themes of global catastrophes however grows, files on risks are made, in some years these ideas will start to get and into authorities of all countries. Means, their defensive value nanotechnologyй is already visible and creation possibility of "grey goo» is clear. The understanding of gravity of risks should сплоить all people on a transition period that they could unite in the face of the general threat.

The analysis of risks of global catastrophes gives us the new point of view on history. Now we can estimate modes and politicians not from the point of view of that good they have made for the country, and from that point of view from which it is visible, how much effectively they prevented global catastrophe. From the point of view of the future inhabitants of XXII century how well or badly we lived will be important not, and how much we have tried in general to live before the future.

In summary it makes sense to express basic unpredictability of global catastrophes. We do not know, whether there will be a global catastrophe, and if yes, how and when. If we could know it «in advance would spread straws». This ignorance is similar to that ignorance which is at each human about time and his death reason (let alone that will be after death), but human has at least an example of other people which gives statistical model of that, as to what probability can occur. At last, though people and not very much like to think of death, but nevertheless from time to time everyone about it thinks and somehow considers in the plans. Scenarios of human extinction are practically superseded in the public unconscious. Global catastrophes are fenced off from us by a veil as the technical ignorance, connected with our ignorance of real orbits of asteroids and to that similar, and psychological, connected with our inability and unwillingness them to predict and analyze. Moreover, global catastrophes are separated from us and theoretical ignorance - we do not know, whether the Artificial intellect, and in what limits is possible, and we do not know how correctly to apply different versions of the theorem of the Doomsday which give absolutely different likelihood estimations of time of a human survival.

We should recognise, that at any level catastrophe has already occurred: the darkness of incomprehensibility shrouding us has eclipsed the clear and clear world of the predicted past. Not without reason one of articles Ника Bostromа is called: «Technological revolutions: Ethics and a policy in darkness». We will need to collect all clearness of consciousness available for us to continue a way to the future.

Part 2. Methodology of the analysis of global risks.