Gaia, Cyborgs, and the Novacene

This should actually have been an interview. James Lovelock’s 2019 book Novacene. The Coming Age of Hyperintelligence contains a succinct presentation of all that could be summed up under the term “digital ecology”. The main thesis of this slim volume, also published in German in 2020, is that humans in the Anthropocene are no longer in a position to do anything substantial to counteract the climate catastrophe; instead, the only ray of hope is the hyperintelligence currently being developed – Lovelock refers to “cyborgs” or “electronic life” – which will possibly find ways and means to stabilise the world’s climate, at least for some regions that will remain habitable. Lovelock does not resolve the question of whether this will happen in time, before most humans fall victim to the overheating largely caused by humankind. The only thing he is certain of is that a new age will dawn, the Novacene (a deliberate pleonasm, since the Greek “kainós”, from which the suffix “-cene” derives, already means “new”). It is, however, more than debatable whether humans will still play a role in this largely digitally determined world of the “new”.
The original plan was to interview the scientist, who celebrated his 102nd birthday in July 2021, to discuss this thesis and its far-reaching implications. The arrangements had already been made and the interview was to have taken place at Lovelock’s country estate in Dorset, England. However, the coronavirus regulations in force at the time (a ten-day quarantine even for vaccinated visitors to Great Britain, and the same again when returning to Austria) made an in-person meeting de facto impossible. Nevertheless, with a view to eliciting some replies, a sporadic e-mail exchange followed, which although not providing more detailed replies to the questions posed, did contain some statements in which Lovelock outlined his lifelong credo of what is known as the Gaia theory and, building on this, the dawning of the Novacene. We can still hope for a more detailed explanation of what would merit designation as “digital ecology” as long as the scientist, born in 1919, continues to enjoy good health. However, given the circumstances, the following cursory remarks will have to suffice here and, of course, raise various questions rather than being explanatory.

Electronic life
The central argument of Novacene is that we are about to enter a new era in which intelligent machines – Lovelock refers to them (somewhat retro-futuristically) as “cyborgs”¹ – will be the next stage in the evolution of life on Earth. As evidence, he cites the rapid increase in computing capacity, which obeys Moore’s famous law (the annual doubling of storage and transmission rates since the mid-1960s, a prime example of exponential growth). It is not simply that machine systems can transfer signals umpteen times faster than the human brain (in practice about ten thousand times faster, as the trained physicist and chemist calculates in a captivatingly succinct manner).² It is also that this forms the main reason to assume that the seeds of a new form of life could also be contained in the accelerated development. It is unclear or highly speculative when that could be the case – the only clue could be that humans “think” about ten thousand times faster than plants due to their genetic blueprint, allowing scope to determine by analogy the timing of the upcoming changing of the guard. However, if the slow development of organic and especially human life is factored in, it might still take a while before fully fledged electronic life takes over.
The background to this, in Lovelock’s view, are the three great scientific revolutions in the Earth’s history: 1. about 3.4 billion years ago, the first photosynthesising organisms (originally bacteria) emerged and began to convert sunlight into energy; 2. the beginnings of the Anthropocene, dated by Lovelock to precisely 1712, when Thomas Newcomen from the UK (a forerunner of James Watt) invented a machine that could convert (solar) energy stored in coal into work; and finally 3. the current threshold to the Novacene, in which sunlight can be converted directly into information, ultimately aiding our cyborg successors to outstrip human life. Lovelock is convinced that this development has already begun, parallel to the Anthropocene, although it only really took off as computer science and cybernetics developed from the 1940s on, making the accelerated evolution we are confronted with today and/or in which we inevitably participate almost uncontrollable.
“In reality, the Novacene, like the Anthropocene, is about engineering. The crucial step that started the Novacene was, I think, the need to use computers to design and make themselves, just as AlphaZero taught itself to play Go.”³ This qualitative leap, which means that artificial intelligence systems are increasingly capable of self-construction or continuous self-modification, represents one of the main aspects of the upcoming epochal transition. Deep Learning, as implemented in neuronal networks for some time, plays a crucial role in this (and can, for example, predict by looking into someone’s eyes how likely it is that they will suffer a heart attack).⁴ However, Lovelock disputes that our conventional causal thinking (an important underpinning of the Anthropocene, derived from Newtonian mechanics) still has any explanatory value in this context. The now increasingly widespread autopoiesis of intelligent systems is situated beyond known mechanical principles (which is perhaps why it is such a dramatic epoch-redefining development).
In an email, Lovelock notes “I think it important to keep in mind that I am more of an inventor than a scientist. In the capacity of an inventor, I find it wrong to refer to a computer or cyborg as mechanical. True enough, Babbage introduced the functioning mechanical computer, but I am fairly sure he would have been much happier to have used electronics instead. The rate at which information is transferred along a neuron in our bodies is 1 million times slower than the speed of travel of an electron along a copper wire. It needed the invention of electron tubes and transistors before electronics could be exploited.
I do not think that a mechanical cyborg would be possible. I once argued with Archer Martin about this. He thought it would be possible to make mechanical motion as rapid as electronics. He was, of course, a genius. Cyborgs do not yet exist unless someone like Demis Hassabis [a British AI researcher] has made them. The million-fold difference in speed of thinking between ourselves and cyborgs does not mean that no relationship is possible. Think of how we view the vegetable kingdom.”⁵

Conditions for survival
Lovelock identifies a further feature of electronic cyborg life in the evolutionary trait of “intentional selection”. In contrast to “natural selection”, which has determined the history of evolution to date, the new life form will ceaselessly optimise and reproduce itself – “errors in these processes are corrected as soon as they are found”.⁶ Purposeful or intentional selection sounds at first like a highly anthropomorphic concept, raising the question at this point, not without good reason, of whether a little “homunculus” (as this stubborn human revenant was once called) does not after all lurk in the background of the brave new Novacene. Or to put it another way: Who or what will determine the intention in question, the purpose of selection? Is it plausible to assume that (human) concepts such as intention and purpose will still matter once electronic life begins to organise and optimise itself? It should be noted though that the history of robotics has always been dominated by the humanoid look, the human-like appearance that led to the renowned uncanny divide between the real and almost-real. What will happen though if we throw this anthropocentric assumption overboard (and the hypostasized self-regulation of cyborg life would at least not exclude this option)?
Lovelock, moreover, harbours no illusions as to the human-like nature of cyborgs. He sees their form dawning more in the shape of spheres, their mode of communication being telepathic, and their mode of movement more in quantum physical teleportation.⁷ “It would obviously be wrong to think of any of the inhabitants of new electronic biospheres as robots or humanoid in any way. They could take the form of a parallel ecosystem ranging from microorganism to animal-sized entities. In other words, this would be another biosphere coexisting with the one we have now. Their natural language would not be the same as ours.”⁸
Even moving away from the long-dominant idea of competition and focusing instead on the coexistence or cooperation so often demanded today, the relationship between humans and machines is complicated, to say the least. In Lovelock’s evolutionary approach, human intelligence will produce a cyborg intelligence that may perhaps (and the “odds” favour this outcome) see no real purpose for humans once this evolution is fully underway. In post-humanist thought there has often been a certain melancholy about this, but Lovelock’s credo moves in a completely different direction: “Do not be depressed by this. We have played our part.”⁹ To sum up the nub of his thesis, that has entailed assisting at the birth of the cyborg intelligence that will ensure the planet’s survival. He is convinced that the Earth’s current state leaves humans and machines no choice but to cooperate across all species barriers, not least to preserve the common planetary habitat that guarantees the survival of cyborgs, humans, and many others.
Perhaps electronic life forms will one day have different requirements for survival than humans, yet, even if that occurs, Lovelock’s benevolent calculation is that they will be concerned first and foremost with the planet’s habitability – or at least with maintaining a habitable zone for their own reproduction. Admittedly, they will at some point be “entirely free of human commands, because they will have evolved from code written by themselves”.¹⁰ That however does not signify that they will therefore necessarily disregard the conditions of their own reproducibility. Rather, this may conceivably hold the seeds of (perhaps the only possible, cross-species) ecological salvation.

“Gaia will keep the peace”
Enter Gaia. Conceived by Lovelock in the mid-1960s and named, picking up on a suggestion from the writer William Golding, after the Greek goddess of the earth, hardly any other concept has had such a pronounced impact on the earth and environmental sciences in the last 50 years. Far removed from any New Age mythology, despite longstanding attempts to deliberately misunderstand the Gaia hypothesis in those terms, Lovelock’s idea of a self-regulating Earth was primarily concerned with a new relationship between biology, geology and atmospheric physics or chemistry. The complex interactions between these three fields had simply not been on the disciplinary agenda previously and were only finally addressed, after decades of hostility, when Lovelock’s theory emerged and was developed further in cooperation with many other scientists, such as biologist Lynn Margulis. Ironically, the idea is now considered common ground in the earth sciences or what is known as geophysiology. However, a long conflict raged over the concept of understanding the biosphere, or the interlocking of various biospheres, as a vast, dynamic, self-regulating system that helps create the conditions in which it can thrive.
Even today, Lovelock is prudent when elaborating on the Gaia idea, about which he has written a good seven books over the last 40 years:¹¹ “I should have explained that self-regulating systems, such as the kitchen oven that keeps its internal temperature constant regardless of changes in the kitchen temperature, or the physiological system that holds our body temperature constant at 37 degrees C or indeed my image of Gaia that holds the Earth’s temperature within a range compatible with life, systems such as these cannot be explained in words.
Newton first discovered this limitation of the written and spoken language. He was obliged to introduce the infinite iteration of calculus. Strong evidence in favour of Gaia as an operating system is the maintenance of life for 3 billion years despite the 20% increase in solar heating”.¹²
As is generally recognised, the risk of warming represents one of the main problems in the now widely acknowledged climate catastrophe that Lovelock was among the first to predict. Nevertheless, he remains rather reserved about the calculation and simulation models deployed for example by the UN Intergovernmental Panel on Climate Change. The mathematics used seems to him to be too linear and one-dimensional, the factors and variables considered not sufficiently complex. “My point is that global warming is certainly real, but the outcomes currently being predicted by scientists, politicians and Greens are not necessarily the ones we should most fear,”¹³ he writes in Novacene. Everything could actually prove to be much more disastrous if, as he has previously pointed out, the temperature trend is not linear but operates in phase jumps, as is often the case with self-regulating systems.¹⁴ Furthermore, the current almost exclusive focus on reducing CO2 emissions does not do justice to the problem as a whole, as for example the concentration of other atmospheric gases, the acidity of the world’s oceans, the spread of seaweed, which is decisive for cloud formation, what is known as the albedo factor [the sunlight reflected by the Earth’s surface], to name just a few factors, would also need to be taken into account to arrive at more reliable temperature forecasts. “Sustainability needs good accounting, and this is where digitisation is most useful. We know that we need to reduce the emission of greenhouse gases (CO2, CH4, and CFCs). The carbon industry should be reduced. It should be regarded as a liability, not a source of wealth. We need time and energy to clean up the mess we have made.”¹⁵

Taking a leap forward from this point to the dawning cyborg era, we might ask to what extent Gaia will still play any role in this scenario. Perhaps the increasingly fine-grained feedback mechanisms on which the self-regulation of future electronic life forms will be based will simply make the idea of one big regulator seem superfluous. That however is where the atmospheric physicist comes in again, noting that even electronic life cannot thrive above 50 degrees Celsius – so henceforth humans and cyborgs should have a common goal, namely “to keep the Earth habitable”. ¹⁶ Irrespective of how hostile or sympathetic the two may be to each other in future in the face of inevitable hierarchical differences, they will always have one thing in common: constraining overheating to guarantee their survival. It is precisely in this sense that it can be said, as Lovelock does with great nonchalance, that Gaia will “keep the peace” ¹⁷ – even if for many species it may already be far too late.

Machines of loving grace
One of the basic ideas that has persisted since the emergence of Anthropocene thinking, indeed in some ways constitutes its core, is that it is already too late to maintain the Earth’s habitability in the long term or on a large scale. The “Age of Fire” is another expression Lovelock uses for the Anthropocene – a term, incidentally, that in his view was coined by American ecologist Eugene Stoermer (rather than, as is generally believed, by atmospheric chemist Paul J. Crutzen). Here, too, Lovelock does not seem as unreservedly pessimistic as he perhaps might in view of the damage already done. Instead, he adopts an eco-modernist stance – a realistic, fact-based perspective that is quite comfortable with geoengineering (if the measures can be reconciled with the bigger picture that needs to be kept in mind). In fact, it is precisely in relation to this larger perspective that AI could play a critical role, since it can cope with data or recursive data processing that no human brain can handle.
In this regard, Lovelock recalls a famous poem by Richard Brautigan – “All Watched Over by Machines of Loving Grace” (1967). In it, the hippie author attempted to fuse romantic back-to-nature idealism with Cold War systems theory based on computers and cybernetics that was all the rage back then. “The idea was that governments and big companies could be eliminated by the creation of a benign cyber system that worked alongside nature”. ¹⁸ As we know, that vision did not really come to fruition, although today, half a century later, we are beginning to catch a glimpse of the promise that collaboration between ecology and cybernetics might hold. In an email, Lovelock outlines the connection as follows: “The term ‘Digital Ecology’ expresses succinctly what Gaia theory is about. Sometime during the 1990s a book on digital ecological modelling was published. Its author was Robert Lord May, who subsequently became President of our Royal Society. The book summarised the numerous attempts to model digitally ecological systems. The book concluded that as Lotka and Volterra found, any attempt to model more than two species was impossible. Lord May himself went further and observed that the mathematics of multi-species systems were prone to chaotic behaviour. Similar breakdown into chaotic behaviour was observed in other multicomponent systems such as Lorenz’s meteorological systems.
In 1991 I submitted a multi-species model that was free of chaos for publication in the Proceedings of the Royal Society. It was rejected by the peer reviewers but shortly afterwards the editor of the journal informed me that it would be published because the President (May) of the Society approved it. The paper is entitled A numerical model for biodiversity (1992). The first paragraph of the paper expresses my debt to Alfred Lotka. He was the first to realise that ecological systems could not be modelled unless the environment is included.”¹⁹
One way or another, cyborgs’ hyperintelligence will be needed if these environmental relationships are to be factored into the equation in a way that chimes with Gaia, rather than stopping at reductive CO2 emission models. Machines can thus prove their redeeming benevolence, even if this only exists in bits and bytes. In general – although this opens up an entirely different issue – Lovelock views the final stage of evolution, the goal of intelligent life, in complete “transformation of the cosmos into information”. ²⁰ It is not by no means certain that organic life or earthly materiality will still play any role at all. Whether Gaia will simply accept all this is probably equally uncertain.