Military Strategy in the Posthuman Era
Image from Kentsuku CG World Copyright
Introductory notes for Virch Course Environment EA3880: History of Interstellar Warfare and Strategy, University of Corona.
The Introduction of AI during the Information Age The introduction of artificial intelligence during the information age was to a significant part driven by the need for more autonomous weapons systems. Information age armies from democratic nations could simply not afford to lose soldiers in battle due to the public outcries at home, and the expense of training personnel for information age warfare made the loss of a single pilot or infantry soldier a noticeable loss. The age of expendable gun fodder was long over.
The first autonomous weapons were support systems for soldiers, integrated into decision support systems, soldier wearable computers or fighter aircraft, but soon independently moving drones or autonomous gun emplacements were growing common. As AI developed further, drone weaponry became used in everything from full-scale military warfare to riot control and security systems.
Perhaps the most important effect was that military agents now could be manufactured rather than recruited. Even if developing and training an AI "soldier" was extremely expensive, the cost of copying the soldier was negligible. This enabled much larger armies than before, although economics and logistics still limited their practical size.
At the same time the infowar side of warfare grew into a parallel battle on the nets and in the ether, where software system battled using encryption, jamming, hacking and denial of service. Having net superiority became just as important as having air superiority, if not more essential. Everywhere AI became essential and more and more planning and execution were assigned to AI systems.
Although the AIs were quickly becoming the de facto military commanders and soldiers in any conflict involved developed groups, military conflicts in the Information and Interplanetary Ages were still fought by sub-singularity entities. The existence of ultra-fast autonomous weapons may have moved human soldiers and decisions into the background, but the lack of information, slow transports, low production rate and primitive weapons did not lead to a significant qualitative change beyond an ultra-fast version of mature information age military doctrines.
Nanowarfare Nanowarfare added a new battleground to the old ground, sea, air, space and information battlefields: the nano theatre. Here human decisions were totally inadequate, making AI support even more essential, but in addition it enabled replicating weapons.
Replicator weaponry, either automated production units producing autonomous weapons and more production units, or self-replicating goos and swarms, turned the old supply paradigms upside down. Previously production capacity had been something static, something that was constructed in peacetime and added new units at a nearly constant rate. With replicators the capacity would instead grow exponentially from the moment it was initiated, making the group that started arming themselves first potentially invincible. Replication could also be started almost anywhere where enough energy and raw material was found.
Replicator weaponry profoundly destabilised the military situation in the solar system, making the Technocalypse nearly unavoidable. As soon as reliable replicators became common enough all sides knew that the only way of avoiding being wiped out by some fraction with their own exponentially expanding power was to start arming themselves. In the early days a situation with two or three sides balancing each other in a situation of mutually assured destruction, holding everybody else back from full replicator capability would perhaps have been possible if the development of nanotechnology had not been such a broad and slow process allowing many sides to participate, the resulting technology had not been pirated and the vastness of the solar system had not precluded easy control. Instead the inherent logic of the situation forced even the most peaceful group to arm itself, if only with massive blue goo. The eventual disaster was a logical consequence.
If replicator weaponry was inherently destabilizing by making preparatory armament so profitable and hence encouraging unlimited arms-races, blue goo instead inherently included evolution into warfare, adding an inherent drive for complexity. Even if virus evolution had been a part of biological and software warfare long before, evolving weaponry to suit the enemy was a new concept in the nano era. Blue goo had to not just adapt to the enemy, but also adapt to its counter-adaptions.
This paradigm was also transferred into macroscale weapons, although large-scale application was delayed until the Federation era. Here field engineering AI was applied to devise new weapons and adaptations on the fly. This to spelled the end of uniform weaponry and units on the same side; through the co-evolutionary races of battle the units evolved into new species as needed. The same occurred in tactics and strategy, and the most successful designs not only survived but also spread to other friendly units. In interstellar warfare the slow communications led to the radical divergence of units.
Hyperturing Doctrines The appearance of hyperturing AI supercharged this adaptability. While early and middle Federation weapons were limited by the ingenuity of sub-singularity AIs, hyperturing AI allowed qualitatively higher forms of intelligence. Hyperturings could on average outsmart the simpler intelligences, making hyperturing forces very effective when striking against subsingularity forces. They were not invincible, as lack of information or resources as well as certain game-theoretical standoffs could still give a victory to the other side, but they had a general advantage.
An typical example of how near baseline forces could deal with hyperturing forces occurred during the Second Consolidation War, when the Geminga Orthodoxy were faced with Metasoft and Dominion forces under hyperturing leadership. Whenever possible they used strategies where they would strike with an entire force against one system of two or more. Which system was attacked was determined using random numbers. This meant the hyperturing opponents either had to divide their forces between the systems, becoming outnumbered in the attacked system, or themselves gamble on which system would be attacked, which would either lead to a fair fight or (more likely) being in the wrong system while the orthodox destroyed a lightly defended system. After each invasion, the orthodox fleet would repeat this strategy, tracing a random walk of destruction through space.
Hyperturings enabled not just enhanced adaptation, strategy and efficiency, they also introduced the concept of stalemating enemies. In some situations hyperturing forces could demonstrate to enemy forces that the only possible outcome was a defeat; there was no reason to fight the battle. In some cases both sides simply self-destructed a number of units corresponding to the minimal predicted mutual losses and retreated in orderly fashions. In others stalemating did not change anything, as a win or defeat on the local level may be part of greater strategies on higher levels, the weaker side not comprehending the full situation or even pretending not to comprehend, in order to bring some tactical surprise into play the apparently superior force may not have taken into account.
To the hyperturing strategists warfare had become an extremely complex game of chess rather than gambling. While this view may have been correct in situations where both sides had plenty of information, fast communications and equal levels of processing, it was still a deceptive trap that could be exploited by clever entities.
For example, during the Battle of Hejne in 4291 the Dominion hyperturing commander of the battleship Digimitim proved to the Jan-Hejne insurgents the inevitability of their defeat. The insurgent AIs feigned a combination of incomprehension and mental rigidity, forcing the Digimitim into battle. At this point they suddenly brought replicated weapons from a nearby moon surface into play, turning the tables on the Dominion. Had the Digimitim attacked directly the insurgents would not have known whether it knew about the weapons or not and been in a far less useful position, but the proof implied that the weapons were not known. In the new situation the Digimitim had to settle for a far less favourable stalemate forcing it out of the system. Commentators have pointed out that in this case the insurgents were lucky that the Digimitim was a fairly well known factor while the exact capabilities and turing-level of the insurgent planners was unknown; had it suspected a trap it might have sent a false proof.
Hyperturing warfare evolved into a Nomic-like game where the rules could be rewritten and winning was based on doing something completely unexpected, something outside the system. This paradoxically gave the low-level units a larger freedom as adding a suitable level of low-level unpredictability often enabled the necessary freedom for higher-level stratagems. On the other hand many downright incomprehensive orders were beamed down from Headquarters that had to be implemented regardless of their apparent madness.
Mature Posthuman Strategy The breaching of further singularity barriers merely extended this complex strategy game into abstraction. To the mainbrains entire wars were fairly comprehensible objects that could be simulated and understood with a high accuracy - while the eventual outcome was somewhat unpredictable statistical modelling and hyperational game theory made it possible to manage them as nearly deterministic chess-pieces. Since the other mainbrains were known to be just as intelligent, the optimal strategies were game-theoretical equilibriums. Rather than winning, the mainbrains sought more elaborate cultural or political goals where the actual warfare was just a minor part.
Compare this to the emerging view that the Version War was not only inevitable from the early Integration onwards, but that the eventual outcome was likely understood among all major powers. Yet they went through with the war, since while the outcome was to a large extent deterministic many important detail-level outcomes were unknown and possible to influence. For example, the eventual fate of many systems and political entities remained indeterminate despite multisingularity level insights into the dynamics. The behaviour of less rational minor polities could not be accurately modelled, and introduced uncertainties that enabled attempts to gain advantages for different empires.
The autowars of the second consolidation war and the version war was the mature form of replicator warfare. Although they were constrained by the limits of autoindustrialisation (early attempts with seeding remote systems with replicator technology showed that they tended to suffer a Denebola-like collapse unless they were directly controlled by significant hyperturing entities, which themselves require an extensive and complex economic infrastructure to survive), the autowars combined both the exponential advantage of replication with adaptive AI and the use of troop buildup in remote and non-visited systems in order to achieve minimal information for hyperturing defensive planning. In fact, the need to keep as much information hidden and unavailable for the enemy encouraged the relativistic warfare, wormhole destruction and spatial extent of the Version war.
In the post-version war era posthuman military strategy in the wormhole nexus has grown fairly stable. The hierarchy of intelligence creates a recursive situation where most levels act to deal with opponents on the same level; it is relatively rare for a high-level entity to get involved in the low-level details. Exceptions are mainly empires such as the Utopia Sphere and Sophic League where the AIs have intense stakes in the welfare of low-level entities, while the strictly hierarchic Solar Dominion instead have a well-defined command structure. Keter and the NoCoZo instead employ a flexible plurarchic or market approach. In the NoCoZo strategy warfare is treated as a market where individual units act as violence or coercion providers, with consultants, adhocracies, networks, microcorps and coordinator teams trading attack offers and strategic analysis in a flexible manner. The result, although formally entirely without singularity boundaries, is actually just as layered as in other empires but the layering is defined by the Law of Comparative Advantage rather than any formal chain of command.
Current Events Three kinds of conflicts have been relevant in the post-ComEmp period and still apply today: empire-empire, empire-minor powers and empires-outside threats. The inter-imperial conflicts are little more than strategic mathematical games, seldom resulting in anything more visible than occasional scored coups and slow changes in policy. The situation appears to have reached a static or very slow-changing stage where few factors upset the balance that has arrived. The Version War and dissolution of the ComEmp appears to have been the last major transitions in the large-scale power structure. The apparent agreement among the mainbrains to limit development of weaponry or other phenomena that could introduce a mutually assured destruction scenario is an example of how they have reached a game-theoretic stable equilibrium.
In conflicts between empires and minor powers, such as the Biovirate - Keter war, the advantage rests solidly on the empires if they can extend their coercive power efficiently. In many cases viable conflicts are based just on the inability of such coercion: the Perseus Rift situation is based on the lack of reliable wormhole routes and massive lack of information about the volume, while the Disarchy is based on a socio-memetic standoff where lack of information, extreme diversity and the introduction of any coercion will immediately produce an opposite reaction making groups align with opposing empires.
External threats pose the greatest problem for the continued stability of the empires. They pose out-of-context problems, entirely unexpected situations on which little information is available and the strategic game cannot be resolved by a simple game theoretic pay-off analysis. The reactions of the imperial mainbrains to developments such as the Paradigm, Amalgamation, Emple-Dokcetics and Laughter Hegemony are varied and express their basic assumptions about the true core distribution of the universe.
In the case of the Paradigm the situation was fairly localized, the only new development was the need of dealing with an aggressive replicating enemy exploiting system-wide autoindustrialisation, something that had not occurred during the Version War. In this case the solution was to pool resources, including with somewhat unlikely sides such as the Corambytia Caretaker and simply scale up anti-nanoswarm measures while exploiting the technology differential between the Paradigm (advanced nanotech) and the imperial fleets (transapientech). The outcome, while uncertain on an individual or system level was predicted to a high accuracy. The Negentropy ultimatum was a logical consequence of the long-range strategies of the Alliance, and while somewhat surprising to sub-singularity beings it was a natural conclusion of the war giving it another buffer volume against the NoCoZo while enabling both an ideological and political resolution.
The case of the Amalgamation is much less clear-cut, as the hegemonizing swarm extends over a vast and unknown volume, is itself at least of third or fourth-level singularity capabilities and fully exploits information denial strategies. Here the war is instead run as the largest application of adaptive warfare ever, using specially created 'prince' hyperturings diverging to meet the demands, while at the same time used to stabilise inner sphere and Perseus politics. The goal of containment is obviously merely temporarily, as the only stable solution is the combined understanding and eradication of the Amalgamation. Note that this poses an interesting challenge since it represents an entity on par with the empires; so far the game-theoretic equilibrium of conflicts among imperial-level intelligences has favoured co-existence, but in this case it appears that the only equilibriums are success or failure - a situation that would be unfamiliar if not unknown to the imperial mainbrains. It may be that the mainbrains will have to choose between developing destabilising higher-order weaponry or trusting the random fluctuations of the front giving them a victory.
In the case of the Emple-Dokcetics the threat is simply lack of information. While the emerging empire may or may not pose a threat to the neighbouring volumes, the real problem is that essential information is lacking about the nature of the M50 disaster, the long-term development of the Emple-Dokcetic culture or for that matter nearly any relevant information about it. This uncertainty doesn't just make interactions and planning for the inevitable contact with the Wormhole Nexus complex and dynamic, but also destabilises the NoCoZo-Sophic-Metasoft balance in the region as the amount of uncertainty reduces the hyperturing strategic inhibitions.
Finally, the Laughter Hegemony represents a young, aggressive minor force that would likely not be a problem with any inner sphere power, but the Sagittarius Transcultural Cooperation remains unclear about the nature of its posthuman noostructure - if it is true that the Cooperation lacks effective hyperturing guidance it would meet the Hegemony on even terms, while if there exists even a discreet hyperturing guidance the situation would be very much in favor of the Cooperation. An interesting possibility is that the Hegemony is also being used as a means to achieve a united noostructure of the otherwise disorganised Cooperation by enterprising entities.
Summary Modern posthuman strategy poses an interesting situation: through the vast information processing capabilities of hyperturings and the limited number of their conflicts become more deterministic and predictable on successively higher complexity layers. At the same time the indeterminacy, uncertainties and chaos that is dominant in the situations when viewed from the subsingularity level are essential for posthuman strategies since they provide the only ways of gaining an advantage beyond the equilibrium. Without the randomness the strategies would simply lock into a game-theoretic equilibrium, while now the goal becomes to exploit these fractal spaces of possibility between the game-theoretical certainties. Lack of information, random events and independent agents are essential for the execution of modern posthuman strategy.
Countering this situation is the escalating co-evolutionary arms-races that occur when adaptive replicating forces meet each other. Here the uncertainty is amplified to macroscopic scales. They can only occur in situations where relevant information is not available or so quickly changing that adaptation does not lock into any equilibrium.