Introduction

Ecology is applied astronomy; yet the past decades of planetary thinking did not do great justice to this fundamental assertion. There is no planetary without the extraplanetary, no environmentalism without the cosmic perspective. In consequence, the grand scales of cosmic, evolutionary time collapse the distinction between biology and technology, as astrobiology and the search for extraterrestrial intelligence (SETI) increasingly recognize how arbitrary it is to base what counts as a legitimate member of the biosphere on the catalog of the terrestrial specimen of folded cosmic matter humans tend to call “life.”

The more one engages with concepts such as biosignatures, technosignatures, or “life-as-we-do-not-know-it,” the more one may believe that these terms just describe distinct facets of the generic communion of agents that populate a wide variety of ecologies hospitable to a plethora of biospheres and genres of intelligences (Cleland 2019; Walker 2024). If that is the case, perhaps it is high time to work toward ecological thinking of a higher order—a truly substrate-agnostic ecology (SAE). This theory works by pruning ecological thinking of the terrestrial biases that mislead humans into treating the question of viable inhabitation of Earth as a problem with sample size N=1. Instead of returning down to Earth (the title of Bruno Latour’s famous manifesto that turned into a slogan of ecological thinking; see Latour 2022), the path to the Long L—the longevity of human planetary community—requires performing a paradoxical gesture of abstraction from the particularities of the home planet of Homo sapiens.

1. The Drake Equation

Every night when you look at the sky, you are looking at the extraplanetary environment that makes terrestrial life possible and that hides cues about the Long L. These cues don’t come across as messages from outer space but as these messages’ apparent lack. This cosmic silence famously bothered Enrico Fermi at a lunch in Los Alamos one day in 1950: Given the vastness of our galaxy, why have we not yet encountered any extraterrestrial forms of intelligence? What started as a scholarly pastime located somewhere between sci-fi and fringe science matured into a systematic effort to specify the conditions of existence of intelligent creatures (also called sophonts). As far as we count ourselves among such species, this necessarily yields implications for our own inhabitation of planet Earth.

This scientific maturation was largely due to Frank Drake, who—a decade after an encounter between gastronomy and astronomy on Fermi’s lunch plate—initiated Project Ozma (Drake 1961). This was the first systematic attempt to intercept potential radio transmissions of extraterrestrial sophonts, putting into practice the idea—simultaneously developed by Giuseppe Cocconi and Philip Morrison—that radio waves may be a viable sign of extraterrestrial intelligence (ETI) (Cocconi and Morrison 1959). At the National Radio Astronomy Observatory in Green Bank, Drake spent four months listening to radio signals from Tau Ceti and Epsilon Eridani—two sun-like stars neighboring our solar system (both just over ten light-years away).

Although Drake did not intercept any alien transmissions, he set SETI’s agenda for the decades to come. In preparation for the first meeting of SETI researchers at Green Bank in 1961, he jotted down an equation designed to estimate the number of radio-communicating sophonts in our galaxy (Drake 1965). With seven parameters—ranging from basic cosmological values to very speculative parameters that included L, the longevity of technologically advanced sophonts—the equation became a litmus test for competing beliefs about the existence of ETI and for research proposals to find signs of their presence among the stars:

N = R_∗∙f_p∙n_e∙f_l∙f_i∙f_c∙L

The main variable (N) represents the total number of planetary communities of communicative extraterrestrials in the galaxy. The first two variables that N is put into relationship with specify the environments in which planets evolve—the galaxy and the solar systems within. Hence, R_* stands for the star-formation rate in the Milky Way, while the fraction of stars with orbiting exoplanets is denoted by f_p. These two values are the least controversial ones: our galaxy is generally estimated to produce around seven new stars per year (with their total mass equivalent to one to two suns), while exoplanet discoveries in the past three decades imply that pretty much every star in the Milky Way is accompanied by an exoplanet population (thus setting the value of f_p close to 1) (Robitaille and Whitney 2010; Cassan et al. 2012).

After these first two parameters, the values become trickier. To specify N, Drake requires us to know the average number of habitable planets per solar system (n_e) and the fraction of habitable planets with life (f_l). The estimates of ne are contingent on how habitability is defined, which depends on our understanding of habitable zone—the orbital region with just the right distance from the host star to heat the planets enough that liquid water can exist on their surfaces but without incinerating the potential organic chemistry bootstrapping there. One recent study sets the value of n_e at 0.38 to 0.88 planet per main-sequence dwarf star (a sun-like star), depending on the definition of a habitable zone (Bryson et al. 2021). However, these numbers apply only to Earth-like, rocky planets with radii ranging from about half to one-and-a-half times Earth’s radius (which puts the most conservative estimates at 287 million terrestrial planets in the Milky Way). Alternative biospheres that can flourish in nonterrestrial conditions (such as those hypothesized to be hiding below the surface of Jupiter’s moon Europa or Saturn’s Titan) are not considered in this definition. Moreover, habitability only guarantees the conditions that enable life’s emergence, not the actual existence of life on a planet. Estimates of f_l are thus a heavily contested terrain, reflecting speculative assumptions about how life originates and how optimistic astrobiologists dare to relate to life’s cosmic diversity and resilience.

Just as f_l depends on the underlying theory of life, the parameter f_i—the fraction of inhabited planets with intelligent life—is involuntarily wedded to underlying assumptions about what intelligence is and whether it necessarily evolves at some stage of the biosphere’s maturation on a given planet. Here, SETI goes into full philosophical mode, since intelligence is a slippery concept, determined by the extent to which we are willing to accept forms of nonhuman intelligence that may significantly differ in their observable signs. With f_c— the fraction of sophont communities with sufficient technological capacities for interstellar communication—things get even worse, since what really guarantees that technology is a necessary trait of every form of intelligence? To Drake, this parameter was mostly about ETI’s ability to send radio signals across long cosmic distances. However, given that humans have communicated via radio waves for only around a hundred years, the terrestrial history of technology does not offer a very solid baseline for the projections of this variable.

Fermi Paradox

The central tenet of SETI is the Fermi Paradox, based on an anecdote recounting Enrico Fermi’s lunchtime bewilderment at the lack of alien visits to Earth thus far, given the apparent vastness of the universe in terms of both space and time. William Newman and Carl Sagan (1981, 291) add precision to its formulation by rephrasing it as a question:

In a galaxy with ∼10¹¹ stars; with planets apparently abundant and the origin of life seemingly requiring very general cosmic circumstances; with the selective advantage of intelligence and technology obvious and with billions of years available for evolution, should not extraterrestrial intelligence be readily detectable?

The Fermi Paradox has sparked a flurry of creative solutions over the years, sometimes leading to the impression that SETI research is a fringe science that has more in common with science fiction than with academic research. One such solution is the zoo hypothesis: no aliens have visited us because Earth is an exclusion zone intentionally isolated from the rest of the universe for the experimental or entertainment purposes of smarter sophonts (Crawford and Schulze-Makuch 2023).

Milan Ćirković (2018)—one of the most prominent theorists deeply engaged with the Fermi Paradox—has cataloged many of the paradox’s solutions under four main labels:

Solipsist solutions, including the zoo hypothesis, question the apparent absence of observable extraterrestrial intelligence (ETI) by positing that ETI is out there but is impossible to observe due to ETI’s intentions. These intentions range from deliberate withdrawal or camouflage to the ETI simulating the universe with us as the only sophonts inside. These solutions also include religious explanations involving a supreme intelligence, such as a god or gods creating the universe with a single unique species in their image.
Rare earth solutions posit that the emergence of intelligent life is very unlikely due to the irreplicable combination of factors leading to the evolution of terrestrial life-forms. Some solutions in this basket include the historical rareness of Earth—it may simply be one of the first planets in a sparsely populated cosmos where intelligent life evolved. Fellow Antikythera researcher Thomas Moynihan (2025) has a lot to say on this topic.
Neocatastrophist solutions rely on a wide range of likely or unlikely disasters that wipe out intelligent species. The list ranges from extinction events prompted by cosmic factors—such as gamma-ray bursts and comet/asteroid impacts—through self-inflicted nuclear doomsdays, engineered pandemics, or rogue AI, to more fictional disasters, including swarms of automated deadly space probes (von Neumann probes, see Matloff 2022) and predatory species that hide in the dark and swiftly eliminate any competitors (these solutions include the dark forest hypothesis, on which another Antikythera colleague—Bogna M. Konior—has written extensively; see Konior 2025).
Logistic solutions pose limits to the expansion of intelligent life in terms of the extent to which it can produce detectable, unambiguous technosignatures and engage in deep space exploration. These solutions usually rule out grandiose visions of space-faring galactic colonizers or propose cosmic migration patterns that steer technologically advanced sophonts out of our sight (this is one of the solutions favored by Ćirković and Bradbury [2006], who hypothesize the existence of a galactic habitable zone for postbiological communities in the colder outer regions of galaxies, where they can maximize the energy efficiency of computational processes).

Some answers to the Fermi Paradox predate its canonical formulation. For example, Konstantin Tsiolkovsky—a foundational figure in rocket engineering—envisioned that advanced planetary communities would undergo a process of transcendence that makes them unobservable to simple beings constrained by ordinary laws of physics (Ćirković 2018, 197). Even before its inception, the Fermi Paradox thus proved to be a fertile ground for philosophical, cosmological, sociological, and technological projections, and it has cultivated this strong speculative charge to this day. Even if it turns out that the Fermi Paradox does not exist—as some authors believe, including Adam Frank, who points to the massive observational constraints that stand in our way of ETI detection (Fridman 2024)—it taps into the spirit of the intuitions that unfurl from the idea of comparative planetology and hence provides a useful speculative device for the analysis of terrestrial affairs related to planetary habitability.

2. The Long L

The last parameter in the Drake equation—L, the average communicative lifetime of a technological community of sophonts—is an altogether different kind of beast. To talk about average lifetimes means to account for a temporal, historical variable (Ćirković 2018, 225–26). The L-parameter ties together the previous parameters. It starts with habitability, since the longevity of communicative sophont communities is strongly related to them keeping their planets habitable, so that they can (willingly or unwillingly) broadcast signs of their existence to the interstellar realm (Balbi and Milan M. Ćirković 2021). In turn, this parameter is influenced by the likelihood of habitable planets actually being inhabited by any life-forms and of the emergence of intelligent life and technologies. Moreover, habitability itself evolves over time, as the entire habitable zone slowly moves outward over the main-sequence star’s life cycle (Kaltenegger 2017).

The L-parameter presents an imaginary bridge between the science of outer space and ecological concerns on Earth. It turns our attention to the fact that finding ourselves on a habitable planet is not enough—the habitability of the planet must be actively cultivated and preserved so that the community of sophonts can continue flourishing. Together with Lisa Messeri, we may say that the L-parameter represents a gesture of cosmic relation, making us realize that “astronomy does not stand apart from terrestrial pursuits but rather deeply informs understandings about Earth and our modes of being in the world” (Messeri 2017, 327). The reason? As David Dunér puts it: The more a given community of sophonts is capable of steering the biocultural evolution of its technological means, the longer the L-value this community can enjoy. This means that we very quickly arrive at assumptions about the social, cultural, or technological organization of such a community, implying a comparison of these assumptions to the actually observed social, cultural, and technological organization of the only sophonts we have a fairly good knowledge of: ourselves.

However, making assumptions about the biocultural evolution and behavior of sophonts means entering a kind of jungle—a topology of pathways, a landscape of possible histories (future included). The L-parameter opens a multidimensional evolutionary space to be surveyed that is contingent on the different value attributions of all previous parameters. Different theories of life, intelligence, and technology yield different options for maintaining a given planet’s hospitability for intelligent life and thus for meaningfully expanding the L-parameter. Conversely, the more one tries to expand the L-parameter, the more one cares about habitability influenced by long-term, plannable processes that unfold on the scale of evolutionary time, as opposed to fast-paced, short-term contingencies that become practically unobservable from the perspective of deep past and deep future.

To put it differently, the L-parameter indicates the correspondence between the unfolding of intraspecies history and interspecies evolution on a given planet. Short L-values may mean that the species failed to level up its historical game to the evolutionary scale. A long L, on the contrary, may index a species that matched the internal organizational patterns of its collective inhabitation of the planet (mediated via technical scaffolds) to the pace and scale that belongs to biospheric processes, thus achieving a metabolic alignment of sorts. This may mean scaling up in some cases (e.g., scaling up carbon drawdown capacities) while scaling down in other cases (e.g., scaling down the rate of deforestation). SETI’s cosmic perspective, powered by the Drake equation’s framework, repaints global history in planetary terms that correspond to the long durée of Earth-system processes, forcing “the history of the planet, the history of life on the planet, and the history of the globe” to converge (Chakrabarty 2019, 1; see also Dick 2012).

Given these remarks, how can we theorize the hypothetical or actual human efforts to expand the L-parameter of our involuntary planetary cosmopolity (i.e., the communion of biological organisms that belong to this planet)? What are our paths to the Long L? The first step is to orient historically contingent ethical-political belief systems toward an ecological horizon, suggesting that addressing environmental sustainability requires inserting human activities into biogeochemical realities that surpass any cultural articulation. Second, we should follow Dipesh Chakrabarty (2019, 18–21) in championing the concept of habitability as the astrobiological displacement of narrow environmental sustainability: habitability departs from a species-specific conception of intergenerational well-being that underwrites the definition of sustainability, since it focuses on the flourishing of the biosphere as such (humans and any other sophonts included). The astrobiological concept of habitability thus obtains normative traction—it no longer only describes the state of affairs on Earth (and potentially elsewhere) but denotes a desirable goal to be achieved and maintained.

As stated by Chakrabarty (2019, 31), “the planet does not address us in quite the same way as our older categories of earth, world, and globe.” But while he wonders what this new form of interpellation should be called, the standpoint of the Long L clearly privileges a mode of addressing that is decisively biological, geological, and chemical rather than political in the standard sense of the word. The Long L points at the tectonics of politics, a deeper kind of politics rooted in normativity not reducible to individual morals or intraspecies discourses for the collective negotiation of value. Rather, it confronts us with lawlike regularities that are vertically imposed on humans as a species (Chakrabarty 2019, 29). This addressing thus represents a categorically different kind of interpellation that designates a categorically different kind of agency—an agency oriented toward biogeochemical metabolic transactions that happen both within the species and across other species. This includes the totality of the biosphere, with its metabolic tethering to older planetary layers. The temporality in which this species agency meaningfully unfolds matches the pace not of human history but of biological evolution (Chakrabarty 2019, 26–27). For that reason, while the concept of history belongs to the discourse about an intraspecies temporality of collective organization, evolution designates an interspecies temporality of mutual adaptation and—down the line—a collective fine-tuning of parameters of planetary habitability. The maintenance of the planet’s habitability by intelligent life is the principal subroutine of this fine-tuning operation, with the assumption that intelligent life manifests in and through much more than just broadband signal transmissions (as originally assumed by Drake): it manifests in environmental modification at large, on the planetary or even extraplanetary level.

3. A Philosophy of Technosignatures

The reason why SETI treats environmental modification as a manifestation of intelligent life has a lot to do with the long history of different proposals for detectable technosignatures (signs of ETI) that go beyond radio communication: artificial megastructures such as Dyson spheres (hypothetical cosmic infrastructures enveloping stars to maximize utilization of the energy they radiate), clandestine space probes potentially sent to our own solar system, or powerful laser beams intentionally sent in our direction (Dyson 1960; Bracewell 1960; Townes and Schwartz 1961). Many of these hypothetical signs of advanced technologies have to do with energy capture, as the Dyson sphere example shows—after all, there is a clear link between the ability of technological sophonts to capture energy and their ability to transform it into powerful tools of interstellar communication. This intuition is captured well by the Kardashev scale, which divides communities of sophonts into three types based on the amount of energy they utilize:

Type I = 4 x 10¹⁶ W: the energy available on present-day Earth;
Type II = 4 x 1026 W: the energy radiated by the host Main Sequence star;
Type III = 4 x 10³⁶ W: the energy available in the entire galaxy (Kardashev 1964, 219).

With each of these types come different expectations about ETI’s capacities: while Type I may be hard to detect even at a very short cosmic distance (as recent studies of Earth’s own potential technosignatures show; see Sheikh et al. 2025), Type II would already be capable of feats including Dyson spheres, heavy geoengineering, or the construction of fleets of automated probes, and Type III may even be capable of astroengineering (for example, building new planets or stars from scratch) or intentionally kickstarting life’s genesis on other planets.

In an evolutionary context, one can leverage technosignatures to describe and analyze changes on a planetary scale. In this view, planets—just like biological species—evolve and undergo major transitions (a claim that needs to be taken with a grain of salt, since—according to our current knowledge—planets can hardly engage in a version of Darwinian evolution driven by the production of offspring, heredity, or mutations). While biological major transitions involve evolutionary leaps that unlock novel evolutionary trajectories—such as the transition from single-cell to multi-cell organisms, the development of endoskeletal structures, or the emergence of the central nervous system (Rosslenbroich 2014)—planetary major transitions represent the emergence of new planetary layers (Furukawa and Walker 2018), such as the transition from geosphere to biosphere and from biosphere to technosphere. The latter is a new planetary paradigm composed of interlocking biotic and abiotic components (especially humans and human-made technologies) that is characterized by its emergent systemic interconnectedness and self-sustaining behavior (in the sense of propagating its spatially distributed existence over time). This paradigm gradually intensified during the Holocene and eventually overhauled earlier planetary dynamics (Haff 2014).

Peter Haff—one of the major scholars elaborating on the concept of the technosphere—traces the possibility of an ongoing major planetary transition while describing the global extent of the technosphere’s influence as well as its resource appropriation, self-conservation, and autonomous agency. Inspired by Vernadsky’s (2025) concept of the noosphere (which is composed of interlocking mechanisms that channel the cultural biogeochemical energy of humanity), Adam Frank, David Grinspoon, and Sara Walker (2022, 48) theorize this transition as an emergence of planetary intelligence: “the acquisition and application of collective knowledge, operating at a planetary scale, which is integrated into the function of coupled planetary systems.” Following the Kardashev scale’s blueprint, this transition can be rendered in energetic terms too, as Frank does in another study with Marina Alberti and Axel Kleidon that distinguishes five planetary classes:

(Figures 2 and 3)

Class I = atmosphere-less planets in thermodynamic equilibrium (radiating as much energy as they absorb from the host star);
Class II = planets with atmospheres that retain some incoming solar radiation as kinetic/chemical energy that drives climate/geological processes;
Class III = planets with simple biospheres that capture free chemical or solar energy;
Class IV = planets with mature biospheres that have considerable agency in their coupling with older planetary systems (e.g., atmospheres);
Class V = planets “in which the activity of an energy-intensive technological species strongly shapes free energy generation and feedbacks” (Frank et al. 2017, 16–18).

According to the authors, Earth is currently on the threshold between Class IV and Class V, since our planet’s technosphere increases rather than decreases its radiative heating. The outgoing radiation is trapped by the thickening layer of anthropogenic greenhouse gases and fed back into the planetary system, leading to radiative forcing that manifests in a warmer and more volatile global climate (Frank et al. 2017, 18). The high presence of greenhouse gases presents in itself a potential technosignature (Haqq-Misra et al. 2025, 15).

Philosophically speaking, technosignatures are SETI’s entry point to conceptualizing technics: the sets of rules, processes, objects, and their systems (material or abstract) that enable environmental modification and intraspecies (or presumably even interspecies) communication, functioning as an organism’s extended phenotype of sorts. Technics function as scaffolds that drive further cultural evolution within the given species (with possible overflows and overlaps between species neighboring the evolutionary tree, as in the case of Homo sapiens and Homo neanderthalensis), leading to a complex intraspecies structure of communication, cognition, customs, modes of subsistence, economic exchange, or environmental adaptation (Caporael et al. 2014). For example, Bernard Stiegler (1998, 17, 246; 2010, 9) understands technics as expanded memory infrastructure: tertiary retention. Similarly, Peter Sloterdijk (2013) places technics at the heart of what it means to be human—he talks about anthropotechnics, “according to which the human being is a product, and can only be understood (if at all) through analyses of the methods and relations of his production” (Capra 2021, 129).

Scaffolds (or technics) are associated with the evolution of human intelligence (Caporael et al. 2014, 7–8). Hence, to search for ETI, the theory of intelligence must be wedded to the theory of intelligence’s scaffolds, so that one can understand not just the given intelligence’s evolution but also its durability and possible observable symptoms from a cosmic perspective. In this respect, technologies—the artifacts that concretize technics into material or abstract objects—are thought to be the most viable targets in SETI, especially because they are commonly associated with societal progress. The more “technological” a community is—so the logic goes—the more advanced; with SETI, we are presumably looking for very advanced communities of sophonts whose feats are visible across vast swaths of spacetime.

This progressive assumption deserves further scrutiny. First, taking technology as a marker of societal progress is a recent invention. The understanding of progress itself has mutated many times throughout modernity: the progress of morals and culture, scientific-technological progress, economic growth quantifiable through GDP or other metrics, etc. What an advanced community of sophonts may look like is thus a very open question. Second, the perceived link between technology and progress seduced early SETI scholars to engage in bombastic visions of spacefaring, with expansive colonizers taking over empty territories of our galaxy, driven by different motifs: resource extraction, power, a superiority complex, benevolence, life-seeding, etc. It is this expansive assumption—grounded in the idea of technological progress—that originally motivated Fermi’s question: Where are they? Surely, they must have superior technological capacities, so it is rational to assume they can visit us. The absence of any ETI evidence thus itself problematizes these assumptions (although now serving a knock-down argument), proving that theories tethered in intraspecies histories (and the narratives these species tell themselves about these histories) may not be the best blueprints for the construction of theories motivated by the generic, cosmic perspective.

Earth’s Technosignatures

Following Freeman Dyson’s credo that “every search for alien civilization should be planned to give interesting results even when no aliens are discovered” (NASA Technosignatures Workshop Participants 2018), SETI often works as a celestial mirror of Earth. This leads to important knowledge about the larger parameter space of human economic, social, historical, and cultural models as well as the current level of technological advancement of this planetary community of bipedal sophonts. An obvious example is the wide range of Fermi Paradox solutions that take their cues from Earth’s recent or deep history—the looming threat of nuclear conflicts, predatory patterns in human social and economic structures, or mass extinction events such as the Chicxulub impact sixty-five million years ago.

A different kind of celestial mirroring can be found in recent research led by Sofia Sheikh (SETI Institute/Breakthrough Listen Berkeley), which has looked at how easily hypothetical outside observers could detect technosignatures at the Earth-2024 level. As the authors suggest in their report, on the scale of interstellar detectability, the planetary-scale impacts of human activity are rather modest—the most evident artificial modification of the planetary surface is the light emitted by large urban settlements at night, detectable at best from the edge of the Oort Cloud at the very outskirts of our solar system. Any industrial pollutants in the atmosphere are not remotely detectable to observers beyond the sun’s closest stellar neighbor, Proxima Centauri (four light-years) (Sheikh et al. 2025).

Radio or laser transmissions thus remain the best Earth-2024 technosignatures, but they also come with a caveat. If we think in the spirit of Woodruff T. Sullivan III et al.’s 1978 paper about the radio signature of our planet, based on standard TV or radio (and more recently mobile broadband) signals, this “leakage” of signals is detectable at a distance of five to six light-years at best, according to Sheikh and colleagues, which is roughly the distance to Barnard’s Star (second closest to the sun after Proxima Centauri). The only outlier would be a solar system communication network such as NASA’s Deep Space Network (DSN), whose signals may be detectable from up to sixty-five light-years away. But only intentionally modulated signals sent out specifically for the purpose of messaging extraterrestrial intelligence (METI) would make a real difference by reaching truly deep space. Such signals include the 1974 Arecibo message beamed in the direction of the M13 star cluster by a group of scientists, including Frank Drake, Linda May, and Carl Sagan, or the Japan Space Agency’s numerous attempts, starting with “Greetings to Altair” in 1983 (Quast 2021).

4. Becoming Substrate-Agnostic

Luckily, there is another way to approach the technosignatures puzzle: cleanse SETI of as much terrestrial bias as possible. Whether we talk about astroengineering or Dyson spheres, these are extrapolations of past, current, or likely future technologies deployed by the planetary community of humans. Such perspectives tacitly universalize the technical evolution we are familiar with and ramify the concept of technology as an interplanetary constant.

The effort to abstract from terrestrial particularities resembles some fundamental theoretical considerations in archaeology, design, and architecture. These disciplines also frequently grapple with defining artificial structures. Design or architecture struggle with defining their own scope. The more they digest the nonhuman dimension of artifactual production, the more they depart from the parochial emphasis on human intentionality (as the defining distinction between purposeless natural objects and purposeful artifacts), catalyzed by the rapid onboarding of AI as a creative agent on its own terms. Archaeology finds its epistemological and methodological foundations at stake: What counts as reliable evidence of a past human culture? To some extent, archaeology’s position is easier than that of architecture and design, since the discipline explicitly constrains its scope to human cultures. Yet, it must also constantly shrug off potential biases inherent in extrapolations from the known pasts (and presents) to keep enough wiggle room for genuine discoveries. Moderate substrate-agnosticism has thus settled across different scientific domains. When it comes to SETI, it seems natural to jump from moderate to radical, given the lofty parameter space of possible ETIs.

Contemporary conceptualizations of technosignatures hint in this substrate-agnostic direction. At the very minimal level, almost all technosignatures can be deemed prima facie consistent with various organic substrates, so that similar technologies can be produced by different life-forms (although this assumption needs a philosophical case for convergent technological evolution across different planetary and biological contexts). The more challenging—and more rewarding—strategy, however, would be to search for the lowest common denominator of the engineered artifacts or artificially modified environments. Some candidates for these agnostic technosignatures would be complexity metrics such as assembly index (itself a promising agnostic biosignature, see below), anomalous regularities in star or planetary positions and transits, terraformation indexes, or hypothetical zero-knowledge communication (Sharma et al. 2023; Smith and Sinapayen 2024; Haqq-Misra et al. 2022; Zenil et al. 2023). All of these maintain a minimal understanding of the artificial structure (in terms of complexity, behavior, environmental modification, or abstract regularity) while enabling the formulation of potential SETI targets in computationally abstract terms (Lazio et al. 2023).

The problem is that once we liberate technosignature studies from terrestrial biases, the very rationale for maintaining any distinction between the artificial and the natural begins to look quite arbitrary. Consequently, somewhere halfway, SETI’s drive toward substrate-agnostic technosignatures meets the astrobiological conceptions of “life-as-we-do-not-know-it”—attempts to abstract from the familiar, Earth-based specimen of life in order to formulate a larger parameter space of alternative biospheres (such as silicon-based biochemistries, methane worlds, or Hycean worlds) (Lingam and Leyb 2021, 14–22; Bartlett and Wong 2020, 4; Cleland 2019), thus ultimately enlarging the range of possible biosignatures (indices of present or past extraterrestrial life) (Dunér 2018, 50–51). The distinction between technosignatures and biosignatures is after all quite fragile: any technosignature is a biosignature, since it is contingent on the existence of sophonts. Moreover, there may be an overlap between technosignatures produced by different forms of life, meaning that any technosignature is (at least moderately) a substrate-agnostic biosignature.

Further trouble arises with the conceptualization of intelligence. Unless one possesses a reliable, general theory of intelligence, attempts to determine the fraction of life-bearing worlds that evolve native sophonts are rather arbitrary. Of course, one can establish a threshold between intelligent and nonintelligent life. But intelligence is a slippery concept. We all accept that most members of the biosphere are prima facie alive, but assigning intelligence is kind of messy. For many of us, dogs are both alive and intelligent, but ask I. P. Pavlov, and he may tell you something like “their behavior is purely reflexive.” Octopuses have enjoyed massive attention over the last decade—consequently, they’ve been successfully spin-doctored into the society of intelligent creatures. But much of Earth’s megafauna is perceived as being somewhere at the bottom of the ladder of intelligence, the rest of the biosphere notwithstanding. While the circle keeps widening with the mounting scientific evidence, the human remains the (mis)measure of intelligence, whereas in terms of life, there is no doubt a microbe is as alive as the readers of this essay.

Avoiding the anthropocentric/Earth-centric fallacy in defining intelligence would mean going full radical and placing intelligence deep into the evolutionary tree, far below the human level. That seems to be the strategy of Blaise Agüera y Arcas (2025), which leads to intelligence coinciding with predictive powers: intelligence is the capacity to continuously project oneself into existence in an environmentally responsive manner (which entails taking variable external inputs into account). However, this definition lowers the threshold for intelligence to such an extent that it coincides not only with predictive power but with life itself. And if that is the case, there is no principal way to distinguish communicative acts and/or environmental modifications enacted by intelligent forms of life from those mediated by forms of life per se. Hence, there are no technosignatures, just biosignatures all the way up.

In this light, talking about substrate-agnostic technosignatures sounds quite puzzling: such technosignatures do away with any historically contingent theory of intelligence, but the more we know about the widespread distribution of intelligence across biotic and abiotic systems, the harder it is to distinguish intelligence from life. From the other end, as the decades of scholarship on artificial life show, any generous theory of life makes it more difficult to clearly distinguish possible life-like technical objects from “biological” life (Boden 1999). At the same time, generous theories of life seem to be needed to explain possible astrobiological diversity in the universe, including postbiological entities some authors favor as SETI targets (Ćirković and Bradbury 2006).

But hold on: If the boundaries between biology and technology are so porous, why do we not collapse biology into technology, instead of privileging the concept of biosignatures? The first reason is the evolutionary depth of the biosphere. What we casually call “living nature” is a collection of historical kinds sharing a particular time-depth interval. The cohesion of this collection is a question of its active maintenance as a separate planetary layer—it is not a metaphysical monolith that emerges once and then sits on a planet like a rock. The second reason is that while the concept of life seems to operate as a flexible, generic term that recognizes basic features widely shared across diverse kinds of phenomena, intelligence or technology have a much harder time playing such a liberal, inclusive role. They seem to work more as tentative labels that designate the human willingness to recognize certain traits across a particular subsection of species. For the sake of simplicity, recall the image of a widening circle, with its center being the point of view of the labeling cognizer. This does not rule out higher-order, more generous concepts of technology or intelligence—it just signals the different roles these concepts play in conceptual schemas; roles that have a lot to do with grasping the positionality of a given species of cognizers in the internal structure of their biosphere. As before, it is possible to recall the distinction between interspecies and intraspecies discourse: life is a concept of interspecies discourse, and intelligence and technology are relative to a given intraspecies discourse. Consequently, intelligence and technology can be treated as substrate-agnostic phenomena that themselves belong to the basket of life’s manifestations in an expanded version of the substrate-agnostic theory of the biosphere: substrate-agnostic ecology (SAE).

Semiotics of Biosignatures: The Case of K2-18b

As David Dunér (2018) points out, the search for biosignatures—indices of present or past extraterrestrial life (alternative biochemistries included)—is a semiotic operation: it requires an interpretation of certain phenomena as expressions or traces of living organisms. Similarly to how fossils represent traces of extinct organisms in paleontology or carbon isotope ratios represent past climate conditions in paleoclimatology, a microbial fossil in a Martian subsurface sample or complex organic molecules in an exoplanet’s atmosphere can all be traces of alien life. Dunér (2018) also distinguishes biosignatures from the narrower concept of biomarkers, which specifically refer to biomolecules characteristic of particular organisms or organic processes.

In general, biosignatures are dependent on understanding “the physical processes that let us link the signifier with the signified” (Dunér 2018, 50). This means that even the strongest indirect evidence of life on a distant planet will remain ambiguous, since for every biologically produced chemical compound, there is a potential abiotic process that can synthesize the same chemical without any life in sight. The trouble thus boils down to a question at once biological and philosophical, as well as metaphysical and epistemological: How to distinguish life from nonlife, biotic processes from abiotic ones? Cole Mathis from the Santa Fe Institute and Harrison B. Smith from the Earth-Life Science Institute in Tokyo sum up this trouble in a simple observation: to have reliable biosignatures, we need a general theory of life; such a theory—needless to say, an uncontroversial definition of life—is exactly what current (astro)biology lacks (Smith and Mathis 2023). As they put it:

There’s a gut instinct to explain features on other worlds as being due to non-life where possible, because of the intuition that non-life is the “simpler” explanation. But it is only the simpler explanation if we assume that living processes are less likely than not—and this is wholly unconstrained. This may sound surprising at first, given that everything in the universe starts out as being abiotic (presumably). But we do not yet have a way to predict whether a system will tend toward living or non-living states. This is, again, in part a consequence of not having a testable theory of life. Until we have a theory of living systems and non-living systems across environments, we are stuck with asking very specific questions about features of life, and features of non-life in well understood environments. (Smith and Mathis 2023, 4)

In other words, as long as astrobiological research keeps focusing primarily on exoplanetary biomarkers based on the conceptions about a limited range of terrestrial or nonterrestrial biospheres, it will continue running into the same obstacles created by the ambiguity and interpretability of biosignatures.

Take the case of one exoplanet that recently made headlines: K2-18b. This exoplanet represents so-called Hycean worlds: planets with hydrogen-rich atmospheres and planet-wide oceans that are larger than Earth but smaller than Neptune (and are hence called sub-Neptunes). Early reports from 2023 and 2024 already hinted at strong spectral imprints of methane and carbon dioxide in K2-18b’s atmosphere—which are both vital greenhouse gases and frequent byproducts of organic activity—with the three-sigma evidence of dimethyl sulfide (DMS) biomarker making breaking news in April 2025 (Madhusudhan et al. 2025; Madhusudhan 2024). Given that DMS is produced on Earth by marine phytoplankton, its presence in the K2-18b spectra was hailed as the closest we have ever come to the detection of extraterrestrial life. All of this is consistent with the theoretical picture of physical processes in Hycean worlds, yet DMS may also be of abiotic origin: it has been produced abiotically in laboratory conditions and found in interstellar medium and comets such as 67P/Churyumov-Gerasimenko, meaning that K2-18b’s DMS readings can be the result of hypothetical processes of nonorganic nature (Hänni et al. 2024; Sanz-Novo et al. 2025; Reed et al. 2024).

Finding “Life as No One Knows It”

Astrobiologists at NASA use the following definition of life: “Life is a self-sustaining system capable of Darwinian evolution” (NASA 2025). This definition implies an underlying theory of life, which explains how living entities are generated—by specific evolutionary, reproductive, and metabolic processes—and in turn orients our expectations about when, where, and how a living system can be encountered. Now, while the self-sustaining-system part of NASA’s definition is highly generic (which has been praised as the definition’s virtue; see Benner 2010), the Darwinian part remains contentious, since there may be non-Darwinian forms of evolution and there are examples of entities not commonly recognized as life that may fulfill both criteria (such as self-replicating programs from experiments by Blaise Agüera y Arcas and colleagues—although the authors make a strong case to consider these as life, see Agüera y Arcas et al. 2024). In other words, the definition’s underlying theory of life is not general enough. Carol Cleland (2012, 140)—one of the legends of the philosophy of science—concisely articulates the conundrum of life theory versus its signatures by stating: “In order to formulate a truly general theory of living systems we need unfamiliar forms of life, and yet in the absence of such a theory it is unlikely that we will recognize them as such if we encounter them.”

Instead of waiting for a reliable theory of life to pop up, agnostic biosignatures can come to the rescue. NASA’s definition of life can guide us here. Despite being constrained to known mechanisms of Darwinian evolution, this definition points to an important intuition that has the potential to be theorized as a generic feature of life: life takes time to emerge and evolve. Smith and Mathis (2023) concur: presumably, everything in the universe starts as abiotic; the transformation to biotic matter comes later, with planetary environments as hot, spinning incubators. Once life emerges—thus increasing the complexity of its planet—it further evolves, and its progress through the evolutionary space of possibilities is again indexed by the increase in the planet’s complexity. If such a generic feature can be captured mathematically, we would have a good candidate for a biosignature liberated from any underlying theory of life.

One proposed agnostic biosignature relies on measuring the statistical complexity of a time series of data from exoplanet observations using epsilon machines. This approach does away with the need for spatially resolved data, as it simply takes a time series of measurements from numerous wavelengths obtained by an instrument such as an infrared telescope. These inputs are processed by the epsilon machine: an algorithm that generates a compressed model capable of producing a time series statistically equivalent to the input time series (hence the model indexes the compressibility of the input, that is, the minimal algorithm that can reproduce it). The assembly index proposed by Sharma et al. (2023) works similarly: it enables measurement of the “temporal depth” of the measured system (its depth in “assembly space”), explained as the number of steps needed to construct the system. In this logic, the temporal depth (or assembly index) of the hydrogen molecule is lower than the temporal depth of the sun, Jupiter, and Earth, respectively, leading to the claim that in terms of its history (or its depth in assembly space), Earth is the largest object in the known universe. Since assembly theory is not a theory of life but a theory for quantifying universal features of evolved objects, it is an agnostic biosignature (just as statistical complexity is).

5. Substrate-Agnostic Ecology

SAE represents a kind of generic geology of morals that shies away from a metaphysical appeal to the terrestrial provenances of “nature” or “biology” as the measures of ecological standards. Its key topological feature is vertical embedding—older layers condition or determine newer layers, bringing forth the geological metaphor of sedimentation as an approximation of the mechanism of planetary vertical constitution. In SAE, one can think about this vertical embedding in terms of three planetary layers: geosphere, biosphere, and technosphere. Now, assign each a temporal depth index: 0 for geosphere, 1 for biosphere, 2 for technosphere. This temporal depth can also be articulated as an index of the generative entrenchment of the given layer: the degree to which the layer is locked in the planet’s evolutionary history, thus conditioning the integrity of “younger” layers and becoming the inevitable platform on which younger layers must build (Schank and Wimsatt 1986).

Generative entrenchment is a concept borrowed from evolutionary biology, where it describes the degree to which a feature is locked into an organism’s developmental process, conditioning the emergence of all subsequent features. In SAE, the geosphere is the most deeply entrenched layer, providing the nonnegotiable physical-chemical platform for the biosphere. The biosphere, in turn, is entrenched relative to the technosphere, which emerges from and depends on it. However, the analysis of the layers does not necessarily end here: one can go deeper and postulate a layer with an index of say -1, representing the expanded ecological context of the planet’s home solar system. For the sake of simplicity though, let us stick to the three layers postulated above. Importantly, this layered, historical structure establishes a vertical cascade of constraints propagated upward across the strata. Thus, any generic theory of ecological relationships within and across these layers must be centered around the analysis of the temporal depth of these relations. This makes it possible to diagram them in abstraction from the particulate substrate while still representing a theory of habitability for the kind of life that emerges on a given planet (in the case of Earth, Earth-like life); a theory of how the space of possibilities provided by the coevolution of biosphere–technosphere dynamics cast its shadow on the given planetary reality.

Instead of baking arbitrary metaphysical distinctions into the models, SAE treats agency as broadly distributed across ecologies: agents (plus their emergent interactions that stabilize into rule-based relations) are the critical analytical units of this approach. However, studying agents and their interactions may prove to be too difficult in situ, given the usually nonlinear or slow unfolding of evolutionary trajectories (at least relative to the human phenomenology of temporal duration) and the spatial dispersion of different ecological assemblages. To overcome this, computational models and simulations provide essential spatial and historical compressions to help effectively analyze SAEs and thus illustrate how one can see in a truly substrate-agnostic manner. Given their ability to model emergent dynamics, agent-based simulations (ABS) are ideal candidates within this domain, especially those that incorporate robust agentic designs based on state-of-the-art LLMs. A science of SAE powered by these tools can plot topologies of planetary history: their shape, bottlenecks, major transitions, blind alleys, primary drivers, and paths to the Long L.

The differentiation of agents should be an emergent feature of substrate-agnostic models themselves, as these models trace lineages of information propagation across scaffolds of various (bio)geochemical makeup. Even when it comes to societies of sophonts such as humans, these societies cannot be modeled as strictly made up of Homo sapiens only—the social dynamics crucially rest on the ecologies these agents are embedded in, including institutions and infrastructures, thus blending human agents with a manifold of nonhuman agentic components. After all, if ecological thinking is to be substrate-agnostic, takeaways from studies of LLM-based actors in simulated environments are in principle translatable to the domains of real-life societies in the wild; the former are not “just a model” of the latter but a different instantiation of the potentially same ecological pattern. Henceforth, a community of humanoid robots in a training factory and a community of manual workers in an actual factory are functional equivalents at a much more granular level than one would be ready to admit. For example, NVIDIA Isaac GR00T is much more than just a humanoid robot development tool; it is a modeling environment for SAEs (NVIDIA et al. 2025). The same holds for generative agent simulations based on believable proxies of human behavior or collectives of simulated subjects—LLM-based ABS that can be instrumental in simulating the emergence of cultural scaffolds, social rules, and institutions (Park et al. 2023; Lai et al. 2024).

SAE also impacts the consideration of relevant scientific subjects, that is, who or what can be in the position of the knower. Just as with any type of agency, the scientific epistemic agency in SAEs is distributed across humans and nonhumans. The future course of sciences will thus have to deal with the proliferation of nonhuman “collaborators” or “interlocutors” that transform sciences into interspecies discourses proper, thus leveraging a different distribution of cognitive assets across species of agents, as well as their indexing, archiving, or signaling affordances. Science in SAEs means science done with and through these ecologies. With endeavors such as the Cetacean Translation Initiative (or CETI, which attempts to translate the acoustic communication of sperm whales), we may be near the point when in natural sciences, animals that used to be “objects of study” will be more akin to “studying partners” (Rodríguez-Garavito et al. 2025). The same may hold for AI agents, depending on one’s willingness to ascribe a genuine communicative agency to them.

The major consequence of SAE is that technological agents are worth of the same kind of “care” as those agents deemed to be “more natural” in a folk metaphysical framework. This immediately suggests an ethical debt-in-the-making toward nonhumans. Just consider how across Earth, machines of different species take over duties of care related to human life, from mundane domestic tasks such as vacuuming to less trivial operations, such as surgeries or therapeutic consulting. Humanoid robotics is gaining traction. In the decades to come, the rapidly aging global population may find itself interacting more with machines than with humans, just because there simply won’t be enough people of working age to take care of the seniors. In turn, some people surely agree that the things we make require our care too (and may pay us back proportionally to the care we give them). Meanwhile, the wave of breakthroughs in artificial intelligence has prompted many scholars to pay better attention to the status of technological objects in human cultural and moral frameworks. Sadly, the result is a contemporary AI ethics fueled by stories about autonomous cars involved in fatal traffic accidents or chatbots advising kids to do away with their parents so that they can indulge in limitless mutual interaction.

As much as these problems present some urgent questions about the cohabitation of humans and machines on this planet, most are posed in a strangely unidirectional way: they all prompt us to rethink what machines owe us, rather than what we owe them (see, e.g., Université de Montréal 2018). In other words, the technical objects usually raise ethical concerns only insofar as they present a possible threat to human integrity; consequently, the remedy to the pathologies of cohabitation between machinic species and humans lies in strengthening their “alignment” with anthropomorphic values (Russell 2019). We hear much less often about how the increasingly autonomous technologies may challenge these established values (for better or worse) or even the basic assumptions about what counts as morality and agency. Even if the moral agency of machines is recognized, it is tightly linked to the manifestation of traits that make them indistinguishable from humans, such as the presence of recognizable emotions, intentions, and moral reasoning (Llorca Albareda et al. 2023). In the most extreme cases, investment in care for AI agents is even pitted against environmental care as a kind of moral hindrance to the latter (Dorsch et al. 2025).

Against the grain of these assumptions, one may probe the philosophical consequences of thinking about technologies as the offspring of the biosphere; that is, the offspring of life as a planetary phenomenon, stretching way beyond the confines of humanity in both space and time. The value of the technosphere is linked to the biosphere it belongs to, which translates into the duty of care imposed on humans as agents within the planetary biosphere. Humans are called upon by Earth, but this calling is the exact opposite of technoskepticism’s call to arms, so widespread in contemporary critical humanities. If—as science-fiction author Karl Schröder (2011) once said—“any sufficiently advanced technology is indistinguishable from nature,” it follows that the evolving technosphere shall not be pitted against the biosphere as its suffocating blanket but as its vital, valuable—even indispensable—component (Likavčan 2025).

Bitter Lesson and SETI

Rich Sutton’s oft-cited blogpost “The Bitter Lesson” (Sutton 2019) postulates that in the long run, all feats of human intelligence will be overcome by systems leveraging advances in computational power (as per Moore’s Law) and statistical reasoning that search across a vast parameter space in the process of learning tasks from scratch. These include tasks as diverse as playing Go, discovering new protein structures, and recognizing speech and objects. While earlier systems relied on embedding task-specific human knowledge (for example, AlphaGo used the experience of amateur gameplays as the baseline for its policy network), more mature versions were stripped of these anthropomorphic imprints to the benefit of their overall performance (AlphaZero). Based on these observations, Sutton (2019) concedes: “We have to learn the bitter lesson that building in how we think we think does not work in the long run.”

Sutton’s bitter lesson applies to AI engineering and to the future of scientific discovery. Hence, it can also be applied to the search for other forms of intelligence in the universe. The blog post’s conclusion—addressing the search for a scientific understanding of complex mental states—thus resonates with the idea of surrendering the truly generic concept of intelligence to those agents better suited to successfully find and employ it. As he states:

All these [mental states] are part of the arbitrary, intrinsically-complex, outside world. They are not what should be built in, as their complexity is endless; instead we should build only the meta-methods that can find and capture this arbitrary complexity. Essential to these methods is that they can find good approximations, but the search for them should be by our methods, not by us. We want AI agents that can discover like we can, not which contain what we have discovered. Building in our discoveries only makes it harder to see how the discovering process can be done. (Sutton 2019)

Perhaps someone should write a story about how, in the not-so-distant future, an AI agent analyzing data from spectroscopic observations of deep space faces the impossible dilemma of having to persuade a group of bipedal mammals that it has found evidence of ETI somewhere high above their heads. The story could perhaps contain a passage describing how somewhere in latent space a supreme concept of intelligence sits on the throne, bearing an n-dimensional topological structure, an abstract crystal of sorts; a crystal that throws its shadow on the phenomenological corners of human-observed reality, causing these cognitive animals to confuse the shadow with its most brilliant, crystalline source.

6. Navigating the Topology of History

The clear motivation for SAE theory is to include technologies among legitimate ecological actors. It is directly inspired by Haff’s (2014) theorization of the technosphere, which narrates its evolution without human agency as the central causal force. SAE furthers this abstract gesture by separating the technosphere from its other terrestrial specificities, such as the current prevalence of silicon-based computing in the planetary technological stack. In other words, the technosphere should be explained not by the nature of the technological means that constitute this planetary layer but by the generic ends of its emergent dynamics: the transformation of evolutionary older layers via the technosphere’s mediation, observable at the planetary scale. In this spirit, the Frank/Alberti/Kleidon SETI-inspired approach is a great step toward a theory of SAEs. Their proposal privileges the long arc of biosphere–technosphere coevolution while abstracting from the terrestrial biases of what life or technology may mean. In this way, they help us understand structures pertinent to these two planetary layers via an abstract topology of possibilities; a topology that identifies generic dynamics instantiated by the uneasy technosphere–biosphere coupling we are familiar with here on Earth. In turn, this abstract topology enables us to address planetary habitability in a substrate-agnostic, evolutionary-historical framework, hinting at the sequential emergence of planetary layers and the dynamics of their convergence/divergence or alignment/misalignment.

Apart from the vertical segmentation of planets into layers, the topology of history can easily model bottlenecks that punctuate the evolution of new layers. SETI often hypothesizes an existence of some form of the Great Filter—a bottleneck in the topology of planetary histories that occurs at some point in the technosphere–biosphere coevolution, selecting only those technospheric constellations that fit the narrow parameter space for the transition (Hanson 1998). The nature of this Great Filter can be related to energy (in this sense, Haff’s “more energy” option tacitly assumes a Great Filter too: level up or face collapse), population (if you prefer to be Malthusian), or some endogenous breakdown (such as global nuclear warfare). Michael Wong and Stuart Bartlett, for example, envision multiple bottlenecks in the topology of possible planetary histories (see Figure 4), with most technosphere–biosphere trajectories (taken into account from the moment of the emergence of planetary-scale computation, or what they call dataome) concluding in a failed transition from Class IV to Class V planets in the Frank/Alberti/Kleidon classification. In addition, they postulate a small fraction of successfully stabilized biosphere–technosphere couplings in so-called homeostatic awakening scenarios and hypothesize an even smaller fraction as a launchpad to the Type III Kardashev scale (Wong and Bartlett 2022). Failed transitions—the most frequent scenario—are labeled asymptotic burnouts, caused by so-called singularities: “trajectories that are headed toward a state of infinite population and energy usage in a finite amount of time” (Wong and Bartlett 2022, 3–4) akin to the hockey-stick-effect Great Acceleration graphs (Steffen et al. 2015).

The theory of asymptotic burnout implies a solution to the Fermi Paradox: ETIs may be sparse and thus hardly detectable because most ETIs go full extinct before they can establish any strong evidence of their presence. Some solutions go even further, speculating that there may be no detectable ETI around us because everyone is too busy managing the intricacies of their technosphere–biosphere coevolution—galactic expansion or astroengineering feats are simply outside the scope of meaningful resource allocation scenarios. Such is also the consequence of the Sustainability Solution to the Fermi Paradox proposed by Jacob Haqq-Misra and Seth Baum (2009, 49): “The absence of ETI observation can be explained by the possibility that exponential growth is not a sustainable development pattern for intelligent civilizations” (see also Mullan and Haqq-Misra 2019).

These solutions do not yield in themselves any estimates of the value of the L-parameter, nor is that their objective. Nonexpansive ETI can exist on vast timelines without ever producing a reliable technosignature, and if the universe is teeming with short-running sophonts grandiosely torching themselves out of existence every here and there, a short L should not in principle prevent their potential detection (in fact, upon closer look, the galaxy may very well resemble a civilizational graveyard). Instead, these solutions double down on the intuition that drives the philosophy of the Long L: The L-parameter is directly tied to ecological considerations here on Earth, representing one of the biggest possible gestures of cosmic relations.

Fermi Paradox solutions wrap the evolutionary pathway of Earth into the speculative topology of SETI scenarios that necessarily orient the debate on what Haqq-Misra and Baum (2009, 50) call “human civilizational management.” This topology—envisioned as a specific rendering of the space of possibilities—is akin to Hertzsprung-Russell’s diagram (Figure 1): the chart of stars organized by their luminosity and temperature, which simultaneously plots the possible evolutionary trajectories of stars by representing the total space of possible stellar types. This includes the “average” stellar life cycle, visualized as a dense region that diagonally cuts through the diagram (the so-called main sequence). Similarly, one can imagine a variation of the Hertzsprung-Russell diagram for planets, plotting their evolutionary trajectories in various terms: habitability, the emergence of life and intelligence, the presence of complex technospheres, etc.

Such a planetary topology of history then expands the scope of what is usually meant by comparative planetology—the astronomical discipline focused on studying similarities and differences between planets in terms of their atmospheres, geologies, basic chemical compositions, orbital features, and potentially even their biogeochemistries. First coined by George Gamow in 1941, comparative planetology embodies the basic gesture of abstraction that propels any thinking about Earth into a larger topology of planetary arrangements, seeing the case of our planet as a specific solution to the riddle presented by the blind forces that make cosmic matter pack itself into various spherical formations. As he argues, we should think about comparative planetology in analogy with anatomy: just as the anatomy of organisms maps not only the space of possible body plans but also implicitly plots the coordinates of the possible evolutionary and developmental trajectories that lead to these body plans (thus ordering the body plans as sequences of evolutionary steps), comparative planetology must be capable of doing the same in the planetary context (Gamow 1948, 60).

7. Achieving the Long L

The philosophical rationale behind the Long L does not require stating a definite value of the parameter that would be desirable or optimal. The Long L is about a qualitative threshold, a phase transition in the metabolic organization of the technosphere, which becomes isomorphic with the planetary layer that generates technologies as prolonged limbs that operate on the fabric of time and are devised to facilitate the biosphere’s evolutionary trajectory. The consequence? Long L ETIs tend to hide in plain sight because their technosignatures are indistinguishable from biosignatures. This does not mean that technospheres do not exist, only that their successful establishment as a planetary layer involves such a degree of mimicry that it amounts to an ontological correspondence. The Sustainability Solution to the Fermi Paradox paves the way to this implication by suggesting that only sustainably developing ETIs can make it through the Great Filter. Moreover, this suggestion implies that the Great Filter takes place before the transition to Kardashev Type II. For that reason, a Class V planet in the Frank/Alberti/Kleidon classification ultimately loops back to Class IV by reinforcing biospheric resilience via newly emerged technospheric armature. The Long L thus works as an environmental norm that recognizes the maintenance of habitability as the vector of a successful continuation of the technosphere’s evolution, including that of the species that constructed it and its communicable presence. This leads to an important corollary: the communicative lifetime of ETI (as the standard definition of the L-parameter goes) overlaps with the lifetime of ETI as such.

In consequence, terms such as “natural” and “artificial” become quite devoid of meaning: they cannot be used as ethical markers for environmentally beneficial or detrimental interventions. Take the example of climate mitigation. If you were to walk out right now and ask a random person on the street about their honest take on solving climate change, chances are that their answer would range somewhere between “better/more technology” and “live in harmony with nature.” (Of course they may also outright deny the very existence of climate change, but I will let you decide whether to count that as an answer.) Now, say you test their intuitions on the range of supposedly radical climate mitigation efforts piloted in the past years: genetically modifying corals to withstand rising sea temperatures or removing CO2 from the atmosphere by direct carbon capture and storage (CSS). Depending on how “natural” or “artificial” these interventions will sound to them, they will probably score very low or very high on the list of desirable mitigation tools, thus fusing folk metaphysics of nature with moral intuitions.

Looking outward into the cosmic void can help us bypass such an argumentative strategy. SAE’s vantage point does not privilege “artificial” over “natural” and vice versa; instead, it treats the technosphere as a reproductive, evolutionary armature that folds into biosphere over the course of evolutionary time. Far from being a kind of suffocating blanket, technospheres are constructs of certain subdomains of biospheres (just as biospheres are constructs of certain subdomains of geospheres that feature complex chemistries); of those species that develop their intraspecies collective organization toward the capacity to claim agency over the metabolic and evolutionary pathways on the planetary level. The deliberate fortification of this agency then signals the potential maturation of the technosphere by its metabolic profile becoming increasingly isomorphic with the biosphere at large.

Just as the particularities of human biology provide a generative platform for cultural evolution, so the technosphere—the historically accumulated material crust of human intraspecies cultural production—feeds back into the biosphere as its vital, strategic scaffold, enabling two types of scaffolding operations: assisted reproduction and assisted evolution. These two functions are the basic forms of habitability maintenance in SAEs.

Assisted reproduction represents a broad class of processes where the technospheric armature functions as a scaffold that carries the continuous existence of some (1) individual organisms, (2) groups, (3) species, and (4) bio(geo)chemical metabolisms.

The first case (1) is easiest to imagine. Assisted reproduction takes place any time some technospheric armature sustains an organism, whether through help with the provision of life-critical energy or material resources, protection from life-threatening environmental conditions or hostile organisms, offering shelter, and so on. This form of assisted reproduction may be observed across many species, including beavers, ants, humans, primates, or birds. On the contrary, assisted reproduction on the group scale (2) usually involves Stieglerian tertiary retention: mnemotechnics for the intergenerational transmission of, among others, rules, customs, habits, heuristics, recipes, and procedures for environmental or body modification. Note that although this description clearly indicates that this form of assisted reproduction has a lot to do with culture, it is still assumed to be an expression of the biosphere’s self-reproduction in the last instance. Planetary-scale computation infrastructures of platform-based information economies are good contemporary examples of the armatures of such assisted reproduction.

When it comes to the third item on the list, assisted reproduction acquires the flavor of technologies for producing new organisms of the given species. Human IVF is probably the most widespread technospheric armature for assisted reproduction of this kind on planet Earth. Cloning technologies would be another example. Finally, the assisted reproduction of bio(geo)chemical metabolisms abstracts from the organisms (and their populations) as the units to be reproduced, since it focuses on any metabolic process that is sustained by technical means. This can include the hydrological regulation of rivers, carbon capture and storage technologies exploiting reactions that partake in carbon-silicon weathering, solar radiation management, and soil fertility management.

With assisted evolution, we get into a more speculative territory, which abstracts from the short-term benefits that individuals can enjoy as a result of investing in technospheric maintenance of the biosphere. Additionally, thinking about assisted evolution deliberately steers away from those considerations that treat evolution as an untouchable “nature’s project,” with any artificial intervention deemed the spinning up of satanic mills (Sullivan III 2013). This is partially the result of broadening the notion of evolution itself to include technical objects, ecologies, or techno-cultural assemblages: a broadening that retroactively bends the meaning of evolution in a biological context and softens the sharp edges of what counts as an evolving entity and how evolution can be legitimately intervened in.

Cultural evolution—carried on the scaffold of technosphere-embedded tertiary retentions—is a great example of assisted evolution, in the sense that humans play the role of important mediators that propagate memetic content. With the emergence of planetary-scale computation, we can even think about the emergence of an evolving sphere analogous to Earth’s biome: a dataome (Scharf 2018). Hence, far from rendering culture as a realm that resists ontological smoothing into the natural, both terms meet halfway, on the generic ground of SAEs. Similarly, technologies undergo evolution, being subject to forces analogous to terrestrial biological evolution, as widely studied in the Schumpeterian tradition of evolutionary economics that models the dynamics of technological innovation. As with cultural evolution, technological evolution contains the human mediatory element. This, however, comes with a caveat as well: The more artificial intelligence penetrates the innovation cycles—assisting in and potentially even taking over the designing and manufacturing process—the more the technosphere seems to extract patterns of human evolutionary agency, embedding them instead into automated feedback loops. The same holds for cultural evolution: LLMs, image generators, and AI agents assist in the evolution of the dataome as they speed it up. In all these cases—following Michael Wong and colleagues (2022)—we are dealing with the enhancement of planetary habitability, as the overall profile of the forms of lyfe (life in the expanded sense) becomes more diverse and the forms more numerous, thus making the planet superhabitable.

The most important aspect of assisted evolution, however, concerns the ultimate interweaving of technosphere and biosphere via the genetic modification and design of biological species, or hybrid biotic-abiotic systems. As far as evolution is first and foremost the propagation of information patterns that are ceaselessly remolded across their carrying substrates—this remolding being the result of the interplay between the forces of adaptation and the constructive agency of organisms themselves that reshape the surrounding web of life—the agnostic ecological perspective does not hinder the folding of the technosphere into the evolutionary pathways of the biosphere, and in fact it should not, as these two layers are ultimately of the same fundamental provenance. Additionally, extended evolutionary synthesis (EES)—the theoretical paradigm in evolutionary biology that is slowly but inevitably replacing the gene-centered, gradualist, and unilaterally adaptationist Modern Synthesis—postulates a reciprocal causality between environments and organisms and takes into consideration a wider range of information substrates beyond genes (namely drivers of epigenetic, ecological, social, or cultural inheritance) (Laland et al. 2015; Gould and Eldredge 1977). Highlighting this reciprocal relation, EES integrates the insights of niche construction theory, thus paving the way for evolutionary processes to spill over into those domains not standardly rendered as biological (in the strict, terrestrially biased sense of the word) (Laland et al. 2016).

Examples of assisted evolution (in the sense of technospheric aid to the biosphere’s evolutionary unfolding) can be:

Genotype modification: The genetic engineering of organisms such as corals, crops, or even higher-order animals for climate adaptation.
Phenotype modification: The bioprogramming of organisms to modify their phenotype (including appearance or behavior), especially in the case of bacteria or single-cell eukaryotes.
Synthetic evolution: The design of new biological species, biotechnological hybrids, or artificial life.

The vantage point of assisted evolution implies that keeping Earth habitable means way more than the conservation of the conditions for Earth-like life: by embracing the idea of conservation, one categorically fails to understand what the planet as the object of maintenance is. Instead, habitability in SAEs means maintaining and enhancing the planet’s genesity—another term coined by Wong et al. (2022), which captures the creative ability of the planet to always recompose available matter into novel forms of existence. This takes into account care for the technosphere as a critical component of any environmental program for climate change mitigation or adaptation. There is a Nietzschean trope hiding here, as aesthetics gets tethered to ethics and the patient labor of coevolution—the collective fine-tuning of environmental parameters—gains deep political(-ecological) valence. Evolutionary novelty is thus consciously embraced as a meaningful kernel of human agency on the interface between the biosphere and the technosphere. At the same time, it turns into an environmental norm of sorts, one that nods at both the ecological limits and their constructive transcendence. The objective of evolutionary novelty thus can be interpreted as the inauguration of “endless forms most beautiful and most wonderful” (Darwin [1859] 2008, 360) irrespective of their origins, whereas the meaning of “beautiful” and “wonderful” lies in how the species dovetails with the grander schema of the ecological embeddings and planetary layers it partakes in (and cuts across).

Finally, talking about limits versus transcendence, the notion of limit has become one of the central tenets of ecological thinking over the past fifty years. From Limits to Growth (Meadows et al. 1972) to the contemporary degrowth/postgrowth movement (Demaria et al. 2013), different forms of limiting human socioeconomic activity are seen as the most logical conclusion of the appeal to planetary boundaries, whether rendered in terms of carbon budgets, carrying capacity of the environment, or doughnuts (Friedlingstein et al. 2022; Georgescu-Roegen 1975, 372; Raworth 2022). Truth be told, limits are the real thing, but they remain open to interpretation—what one intellectual wing may interpret as an inevitable diminishing of collective agency, the other may very easily see as the constructive precondition for bootstrapping real planetary agency on the species level. Instead of rendering this agency in standard political-ecological terms (which would simply squeeze the exteriority of the cosmic and the planetary into narrow confines of human intraspecies political struggles), one can postulate a kind of environmental normativity here that is not reducible to politics (or ethics) (Likavčan 2022). In this normative dimension, environmental constraints sculpt the space of what is possible: they offer a topology of evolutionary trajectories consistent with prolonging the L-parameter. This presents an important corrective to the value of evolutionary novelty, since it must be measured relative to its constructive contribution to keeping a planet habitable.

Seager Equation

The biosignature–technosignature collapse (with its philosophical reverberations expressed by the biosphere–technosphere ontological lamination) takes its toll on the Drake equation. In particular, all parameters that prove difficult to posit in a sufficiently substrate-agnostic manner can be rolled into the last parameter of the most generic instance: f_l (the fraction of habitable planets on which some form of life evolves). Fortunately, this does not rule out increasing the granularity of the parameters that precede f_l in the equation’s order. One major proposal for the revision of the Drake equation consistent with this line of reasoning comes from exoplanet astronomer Sara Seager. The Seager equation reads as follows:

N=N^*f_qf_hzf_of_lf_s

In Seager’s version, N denotes the number of planets with detectable signs of life, not the number of ETI planetary communities. Consequently, the whole equation acquires a very different meaning—it focuses on the search for detectable biospheres. The parameter N^* stands here for the number of observed stars, while f_^q stands for the fraction of quiet stars, namely those that do not exhibit persistently volatile behavior leading to frequent solar flares or extreme space weather in their neighborhood. Further, f_^hz is the fraction of stars with rocky planets in the habitable zone, no matter whether they are terrestrial planets or super-Earths/sub-Neptunes. Parameter f_^o denotes the fraction of rocky planets in the habitable zone that can be observed, f_^l is the fraction of actually inhabited planets of this kind, and, finally, f_^s is the fraction of these planets where the biosphere produces a detectable signature gas (NASA 2013).

What is the virtue of the Seager equation? It takes into account observational constraints currently in place and thus works as a pragmatic tool for directing astrobiological research into those areas that—all things considered—seem to yield the most promising candidates for life detection. For this reason, it is not a substrate-agnostic equation, although it does implicitly reduce the problem of SETI to the problem of the search for forms of biosignature-producing biospheres capable of inhabiting rocky exoplanets. To increase its substrate-agnostic qualities, it would have to account for forms of biosignatures other than detectable (atmospheric) gases; otherwise, it successfully demonstrates how the Drake equation can become a platform for more targeted attempts at extraterrestrial life detection.

8. Philosophy after Long L

Like many things originating in astronomy, the Long L has a lot to do with optics. Historically, astronomy has been an optical science reliant on the observation of electromagnetic wavelengths across the whole spectrum. A philosophy of the Long L is a kind of optical device, a set of lenses that renders planetary history in a generic parameter space of habitability maintenance by sophont communities. Such a topology of history unfolds in evolutionary time (the native temporal resolution of SAE), which furthers the gesture of cosmic abstraction by accounting for alternative biospheres and technospheres far beyond earthly provenances, parrying myopic conceptions of life, technology, and intelligence. The job of SAE is to generate new philosophical optics on these phenomena. In this respect, the Long L hints at the deep utility of philosophy itself as an optical discipline of sorts: the science of conceptions, theories, and metaphors that calibrate access to the real. After all, the philosophers of early modernity held optics in high esteem: Thomas Hobbes and René Descartes wrote volumes on bending and reflecting light, Gottfried Wilhelm Leibniz placed optical metaphors at the heart of his metaphysics, and Baruch Spinoza was a lens grinder. Today, SAE mobilizes philosophy as an exercise in intellectual lens grinding.

Philosophy belongs to those sciences that are oriented toward a human audience: it narrates certain interspecies realities in renderings charitable to the cognitive dispositions of the human species, relying on metaphor, analogy, and conceptual abstraction. From a metaphysical perspective, thinking about the Long L then functions as a litmus test for some volatile concepts, supposedly constructed to enhance human orientation in the world. For example, technology proves to be a concept that is both recent in its origin and unstable in its content, suggesting that in the long run, it denotes something transitory rather than permanent. From the perspective of SAEs, the power of some concepts may fade away (including artificial, natural), while some concepts may prove to be vital to whatever paradigms come next (most notably the concept of ecology itself).

Regardless of which concepts will survive and which will not, philosophy will retain its orientational character as a science cultivating linguistic capacities that render intuitions about the external world that anchor potential answers to basic questions meaningful and communicable. These questions include how we can know something (epistemology) and what our inter- and intraspecies obligations are (ethics, political theory). The role of language as a vehicle of philosophical conception also points to the general necessity to treat language as a powerful (although not exclusive) scaffold of cultural evolution—language enables intelligence-scaling at relatively low energy costs and provides a safe and expressive “simulation environment” to learn, probe, and question the way the world is without heavily investing in its immediate material transformation. It enables the efficient construction of complexity through which intelligence can scale up.

Philosophy has its place in cultural evolution as one of the less obvious engines of genesity. The collective fine-tuning of the parameters of habitability means building robust, redundant, resilient systems that incorporate evolutionary novelty into endogenous feedback loops. More often than not, this novelty generates a widening gap between the intuitions of human cognizers about the ecologies that surround them and the actual reality of these ecologies. The role of philosophy is not to reduce this gap by forcing the reality of these ecologies into the narrow confines of first-person phenomenology but to productively elaborate on it, to cash out from the opportunities for conceptual engineering this gap opens. In the case of SAEs, catering to the gap between the real and the intuitable means for philosophy to double down on the generic metaphysics of agents, interactions, and planetary layers against the tendency to roll the language back into the differentiation between the biological and the technological as well as the cosmic and the terrestrial. Instead, let us see the latter as a special case of the former. SAE aims to be a philosophical gateway to reasserting the counterintuitive but real aspects of what “ecology” means in its Greek root oikos logos (“a study of home” or “an order of dwelling place”) (cf. Haeckel 1866, 286; see also Watts et al. 2019, 682). This is an order or knowledge not about one particular kind of place but about those generic relations that make those places known as “planets”: homes to uncountable forms of life.

Why Ecology?

In the nineteenth century, the Darwinian paradigm shift in biology codified evolution by natural selection as the fundamental principle through which the characteristics of biological species can be understood. One of Darwin’s key messages was that adaptation to environments was paramount to an organism’s ability to survive, reproduce, and effectively compete with other species in the same environment. This tight coupling between an organism and its surroundings was already observed by the previous generation of naturalists, such as Alexander von Humboldt, but a sufficient theory of the mechanism according to which this coupling functioned was still lacking (Richards 2009, 21). Darwin’s formulation of the principles of natural selection shed light on this mechanism, encouraging further studies of the relationships between organisms and their environments to specify how different species evolve in tandem with their environment (and each other). The leading promoter of these studies, Ernst Haeckel, became one of Darwin’s disciples. In his early magnum opus General Morphology of Organisms (1866), he used the insights of evolutionary theory to retell the history of life on Earth as a story of environmental adaptation (Richards 2009, 8). This book also contains his definition of ecology, a term he coined to describe “the entire science of the relationships of the organism to its surrounding external world.”

Haeckel’s notion of ecology establishes, through evolutionary theory, the essential continuum between the existential conditions of biological species (which, according to Haeckel, are of both organic and inorganic nature) and the history of the environments to which these species become adapted (alongside the history of interactions with other species in the same environment) (Schwartz and Jax 2011, 150). The term ecology thus captures the dynamics of agent–environment interactions and their mutual shaping, which later unfolds into the concept of niche by the generation of ecologists following Haeckel’s pioneering work, including Joseph Grinnell and Charles Elton. Niche denotes “a set of properties of an environment, to which an organismal form may fit” (Walsh 2015, 169). The aspect of the mutual shaping—the dialectical interplay between organisms and environments—is further captured by the concept of niche construction, which replaces the adaptationist metaphor of an organism fitting into the environment with a more dynamic picture of the coevolution of organisms and environments as active constructors of each other’s affordances. As Richard Lewontin (2001, 64) puts it:

What is left out of this adaptive description of organism and environment is the fact, clear to all natural historians, that the environments of organisms are made by the organisms themselves as a consequence of their own life activities. . . . Organisms do not adapt to their environments; they construct them out of the bits and pieces of the external world.

It is exactly this postulation of the constructive relationship between agents and environments that makes the concept of ecology a powerful candidate for a substrate-agnostic framework on habitability and the evolution of life, technology, and intelligence from a planetary and extraplanetary perspective—all one needs to capture this evolution is a generic model that contains categories of agents, environments, and relationships (both agent-to-environment and agent-to-agent relationships). Such a model does not intrinsically rely on a particular substrate of evolutionary change, such as DNA as a biochemical basis of heredity in terrestrial organisms (note that evolutionary theory was formulated well before the discovery of DNA). Hence, it can be compatible with any agents capable of any form of evolution, and any (planetary) environment capable of generating and hosting evolvable agents. Once we are willing to accept technological agents as elements within the biosphere—as products of the long history of agent–environment co-construction—there are no principal obstacles (apart from metaphysical superstitions) to thinking about technologies as ecological agents proper that hence naturally participate in the maintenance of planetary habitability.