Given the pervasiveness of epistasis, adaptation via changes in genetic makeup becomes primarily a search for coadapted sets of alleles–alleles of different genes which together significantly augment the performance of the corresponding phenotype. It should be clear that coadaptation depends strongly upon the environment of the phenotype. The large coadapted set of alleles which produce gills in fish augments performance only in aquatic environments. This dependence of coadaptation upon characteristics of the environment gives rise to the notion of an environmental niche, taken here to mean a set of features of the environment which can be exploited by an appropriate organization of the phenotype. (This is a broader interpretation than the usual one which limits niche to those environmental features particularly exploited by a given species.) Examples of environmental niches fitting this interpretation are: (i) an oxygen-poor, sulfur-rich environment such as is found at the bottom of ponds with large amounts of decaying matter–a class of anaerobic bacteria, the thiobacilli, exploits this niche by means of a complex of enzymes enabling them to use sulfur in place of oxygen to carry out oxidation; (ii) the “bee-rich” environment exploited by the orchid Ophrys apifera which has a flower mimicking the bee closely enough to induce pollination via attempted copulation by the male bees; (iii) the environment rich in atmospheric vibrations in the frequency range of 50 to 50,000 cycles per second – the bones of the mammalian ear are a particular adaptation of parts of the reptilian jaw which aids in the detection of these vibrations, an adaptation which clearly must be coordinated with many other adaptations, including a sophisticated information-processing network, before it can improve an organism’s chances of survival. It is important to note that quite distinct coadapted sets of alleles can exploit the same environmental niche. Thus, the eye of aquatic mammals and the (functionally similar) eye of the octopus exploit the same environmental niche, but are due to coadapted sets of alleles of entirely unrelated sets of genes. (iv) the environment rich in depressive emotion – the aesthetic of Neon Genesis Evangelion are a particular adaptation in qualia-space which aids in the detection/exploitation of the depressive environment.
The various environmental niches E ∈ ε define different opportunities for adaptation open to the genetic system. To exploit these opportunities, the genetic system must select and use the sets of coadapted alleles which produce the appropriate phenotypic characteristics. The central question for genetic systems is: How are initially unsuited structures transformed to an observed range of structures suited to a variety of environmental niches ε? To attempt a general answer to this question, we need a well-developed formal framework. The framework available at this point is insufficient, even for a careful description of a candidate adaptive plan τ for genetic systems, unlike the case of the simpler artificial system. A fortiori, questions about such adaptive plans, and critical questions about efficiency, must wait upon further development of the framework. We can explore here some of the requirements an adaptive plan τ must meet if it is to be relevant to data about genetics and evolution.
In beginning this exploration we can make good use of a concept from mathematical genetics. The action of the environment E ∈ ε upon the phenotype (and thereby upon the genotype A ∈ α) is typically summarized in mathematical studies of genetics by a single performance measure μ called fitness. Roughly, the fitness of a phenotype is the number of its offspring which survive to reproduce. This measure rests upon a universal, and familiar, feature of biological systems: Every individual (phenotype) exists as a member of a population of similar individuals, a population constantly in flux because of the reproduction and death of the individuals comprising it. The fitness of an individual is clearly related to its influence upon the future development of the population. When many offspring of a given individual survive to reproduce, then many members of the resulting population, the “next generation,” will carry the alleles of that individual. Genotypes and phenotypes of the next generation will be influenced accordingly. This is especially important in light of a big universe. If we assume that consciousness is not epiphenomenal, but instead described fully as a slice in the causality of Platonia, then understanding the fitness of degraded experiences barely holding above water by the grace of quantum immortality becomes important.
Fitness, viewed as a measure of the genotype’s influence upon the future, introduces a concept useful through the whole spectrum of adaptation. A good way to see this concept in wider context is to view the testing of genotypes as a sampling procedure. The sample space in this case is the set of all genotypes α and the outcome of each sample is the performance μ of the corresponding phenotype. The general question associated with fitness, then, is: To what extent does the outcome μ(A) of a sample A ∈ α influence or alter the sampling plan τ (the kinds of samples to be taken in the future)? Looking backward instead of forward, we encounter a closely related question: How does the history of the outcomes of previous samples influence the current sampling plan? The answers to these questions go far toward determining the basic character of any adaptive process. But the question is incredibly complicated when we want to measure fitness of experiences, which necessarily exist in an eternal object, and are themselves eternal. How can bounds even be drawn on them?
The answer to the first question, for genetic systems, is that the future influence of each individual A ∈ α is directly proportional to the sampled performance μ(A). This relation need not be so in general – there are many well-established procedures for optimization, inference, mathematical learning, etc., where the relation between sampled performance and future sampling is quite different. Nevertheless, reproduction in proportion to measured performance is an important concept which can be generalized to yield sampling plans – reproductive plans – applicable to any adaptive problem (including the broad class of problems where there is no natural notion of reproduction). Moreover, once reproductive plans have been defined in the formal framework, it can be proved that they are efficient (in a reasonable sense) over a very broad range of conditions.
A part of the answer to the second question, for genetic systems, comes from the observation that future populations can only develop via reproduction of individuals in the current population. Whatever history is retained must be represented in the current population. In particular, the population must serve as a summary of observed sample values (performances). The population thereby has the same relation to an adaptive process that the notion of (complete) state has to the laws of physics or the transition functions of automata theory. Knowing the population structure or state enables one to determine the future without any additional information about the past of the system. (That is, different sampling sequences which arrive at the same population will have exactly the same influence on the future.) The state concept has been used as a foundation stone for formal models in a wide variety of fields.
An understanding of the two questions just posed leads to a deeper understanding of the requirements on a genetic adaptive plan. It also leads to an apparent dilemma. On the one hand, if offspring are simple duplicates of fit members of the population, fitness is preserved but there is no provision for improvement. On the other hand, letting offspring be produced by simple random variation (a process practically identical to enumeration) yields a maximum of new variants but makes no provision for retention of advances already made. The dilemma is sharpened like a fine chef’s sushi blade by two biological facts: (1) In biological populations consisting of advanced organisms (say vertebrates) no two individuals possess identical chromosomes (barring identical twins and the like). This is so even if we look over many (all) successive generations. (2) In realistic cases, the overwhelming proportion of possible variants (all possible allele combinations, not just those observed) are incapable of surviving to produce offspring in the environments encountered. Thus, by observation (1), advances in fitness are not retained by simple duplication. At the same time, by observation (2), the observed lack of identity cannot result from simple random variation.
As Karl Popper observed (before changing his mind eventually, to be fair): natural selection is generalizable to everything: the cosmos, biology, cultural ideas. However, it is my contention that its explanatory power breaks down when considering the competition between Moloch consciousness (i.e. self-aware processes in humanity, transhumanity, and all other arbitrary organisms and AIs across the multiverse) and simple consciousness (that range of most simple experience – whether that ends up being Quantum Torment-flavored or something like unity with Brahman). In other words, once computational specificity/complexity degrades past a certain point, it is unclear how anything is differentially “reborn” since degradation of specificity involves becoming an identical configuration to many “others” (and hence not other in any strictly meaningful sense). The action of the environment upon the phenotype seems to slip past some kind of event horizon.