The evidence that neither theory got it right
The Gaia hypothesis claims that ecosystems self-regulate and stabilize themselves. The discovery of ecosystem collapses caused simply by instabilities in ecological interaction, and that environmental changes in geological history disproves the idea that nature works to preserve an "optimum state", led to the creation of the Medea hypothesis, in which life destroys itself through maximum reproduction without limits (reference: The Medea hypothesis: When life destroys itself). And yet there is evidence that, although Medean events have happened, they are much rarer than they should be according to Malthusian theory (reference: It can be easily shown that Malthusian theory predicts ecosystem collapse in days at best and hours at worst when fast-breeding microbial (microscopic) life is taken into account, and even restricted to relatively fast-breeding animal life the collapse would come in less than a decade). So obviously there is something non-Medean going on, although it is not infallable. Add ImageThere is also empirical evidence that in most cases, bear males that were supposedly going to kill cubs actually loses interest and goes feeding elsewhere instead.
The Gaia hypothesis appears to be supported by simple computer simulations of ecosystems, which stabilize themselves. But when the simulations are made more complex, they start collapsing incessantly and resembling the Medea hypothesis. But real life is so complex, far more so than any of the computer simulations, that theoretically no life whatsoever should be able to survive even a few years of the chaos. For example, predators overkilling their prey and completely exterminating some ecologically important species days before starving to death themselves. This is documented in the case of recently introduced organisms in new places. It have been claimed (by the Gaia hypothesis) that there should be a long-term stabilization, but the complex simulations (which agree with the Medea hypothesis) does not support any such stabilization in any way.
How can it get that way?
But what is causing the non-Medean effect? The Medeanist argument that "nature is not teleological" or "nature does not act on purpose" is just hiding behind the lack of a clear definition of the words "teleology" and "purpose". Any claims of purpose being a "magical ingredient" separate from physics is actually dualism, exactly what the selfish gene proponents who support the Medea hypothesis claims to oppose. From a non-dualist, ignostic perspective, it becomes clear that the whole division into "purposeful" and "purposeless" is just a meaningless semantic play with words. Exactly how much emergent effect-to-cause feedback that comes from the physical processes of ecosystem interactions is unknown, but the orders of magnitude of missing frequency of Medean events proves that something completely different from maximization of the amount of offspring must be driving evolution. If offspring-maximizing evolution were capable of imposing any fixed limits to the flexibility of life, all the worst expectations of Malthusian theory would have come true. So obviously offspring-maximizing evolution is not capable of imposing any fixed limits to the flexibility of living things.
Some form of planning for the future actually follows inevitably from any form of associative learning. Any association between different things, events or whatever effectively means anticipation of one whenever the other is present. This means that foresight is not unique to humans or even higher animals like apes and cetaceans, but is instead present in anything with any association, even amoebas. The objection about life not being able to cooperate to prevent ecosystem collapses totally ignores the evidence for altruism unexplainable by selfish gene theory. This form of foresight does NOT require any thoughts about self identity. In fact, it can be helped by lack of self-awareness, since not being self aware hampers egoistic ignoring of others. See also what is self-awareness? and the exploiticism/dogmaticism fallacy.
"Invasive species" wreaking havoc to ecosystems can be explained by the fact that they have not yet learned from their mistakes about their new environment. For how the insight is retained, see inheritance of acquired characteristics.
There are computer simulations that appear to support the selfish gene model, but they are simple simulations. The most likely reason why the worst Malthusian expectations are not coming true in real life is that the more complex environments in real life is inimical to such unselfcritical entities as the selfish gene model claims that living things are. The strong Gaia hypothesis also appears to be supported by simple computer simulations. So the Medeanists (read selfish gene-ists) made the same methodological mistake as the strong Gaianists, namely uncritically scaling up simple simulations to the greater complexity of real life.
As shown in self-organization, life is actually defined by self-correction, so the fact that planets do not reproduce is totally irrelevant to the question of whether or not a whole planet can be considered a single living organism. The question is really about whether or not the information flow is enough.
But what about destruction of the environment?
So why destruction of the environment? As shown in Brain, plasticity requires a tolerant environment. Social pressure to justify leads to justification which paralyzes self-correction and pressure to blame on others leads to making up of putative mental limitations. So-called "human nature" is just such fake limits, see also Self-organization and Inheritance of acquired characteristics.
So the solution is to eradicate all social pressure to blame on others and justify, see advice of ways to stop justifying. The danger of justification must become widely known. To avoid prestige-based justification, NEVER ridicule anyone for what ideas they came up with. Always stress that nobody can be expected to be right all the time and that only evidence can tell what is right, see also How to avoid hoaxes in science. If no better solution is known, brutality may be necessary in survival situations. But pressure to justify the brutality risks to conserve the brutality even when it is no longer necessary, as history also shows. So justification is the root of all true evil. Since threat of punishment creates pressure to justify and blame away, punitive enforcement is counterproductive and unsustainable. When it gets so severe that almost everyone must committ crimes to survive, the risk of getting caught will be minimal if everybody confesses what they have done, and when all jails are full there will be nothing to threat the starving masses to pay tax to pay the jails, so the few who are caught will soon be free (consider that if children were forced to defend their actions and keep doing so as they grew up they would remain cruel and that ages of liability are arbitrary cultural constructs, and also that the first moral individual evolvability paradox outlined in brain as well as historical evidence for the dangers of vengeance and justification shows the necessity of forgiving past actions). By then there will be lots of Disaster cooperation that puts an end to egoistic short-term thinking. And then, the historical lessons of the danger of blaming and justifying must NEVER EVER be forgot, and the stupidity pattern will never reappear. Then intellectual Gaia can permanently trump antiintellectual Medea.
The counterproductivity of punishment
Research on animal training clearly shows that punishment is counterproductive, and yet animals punish each other. Why? Natural selection cannot possibly have selected for something counterproductive. This means that punishment emerged without natural selection for it. The counter-productivity of punishment means that the existence of punishment/vengeance is due to the fact that the punisher thinks it works (timeservers create an illusion of obedience), consider the fact that violence rates are highest in countries with severe punishment and lowest in countries with mild punishment, implying that a significant factor in the decrease of violence is that punishments have become less severe which falsifies any claims that penal law should be a protection. Any sign of inheritance of a will to punish can be explained by inheritance of acquired characteristics. Because punishment is counter-productive, there is no way natural selection could have selected for it. Indeed, it should have selected against it. The reason why natural selection failed to eliminate punishment is most likely because natural selection by mere death or sterility of whole individuals is a way too blunt means of evolution to have any chance to fine-tune complex behavior, see self-organization.
Any attempts to explain the counterproductivity of punishment by Darwinian selection for alliances clashes with the fundamental Darwinian assumption of random mutations. If mutations were truly random, then evolution of a new instinct would always have to take into account that there would always have to be a one first individual with the new instinct and its viability where no other individuals shared that new instinct. In the absence of other individuals reacting against punishment, a first individual reacting against punishment would not have survived. So whatever the reason why punishment is counterproductive is, it cannot be Darwinian. And yet there is evidence that punishment is counterproductive.
What to do about it
Before the collapse comes, there are ways to prepare for rationality. One is to use foreign languages, which have been shown to eliminate biases such as the otherwise common "losing something is worse than winning the same amount is good" thing. Another way is to avoid ridiculing each other for past or novel ideas in day-to-day life. Spreading knowledge about the inheritance of acquired characteristics is an important way to eliminate the false dichotomy between rigid curriculums and giving up hope, see brain. Practising the scientific method outside the antiintellectual Machiavellism of the official "scientific community", such as on Pure Science Wiki (or other comparable sites) is another way.
Preventing ecological collapse through geoengineering would create possibilities to abuse the geoengineering technology as weapons. And since the counterproductivity of punishment means that the "war on terror" is counterproductive as well, a rational world will still be necessary to prevent apocalypse.