AI is the most important cultural adaptation since the invention of language

In Bacteria to AI, N. Katherine Hayles explores how humanity’s future is intertwined with nonhuman entities — from bacteria to artificial intelligence. Published by First Edition in 2025, the book challenges anthropocentric thinking and introduces a powerful new cognitive framework that includes both biological and artificial agents.

Hello N. Katherine Hayles, why did you write this book, and why now?

N. Katherine Hayles: I wrote Bacteria to AI: Human Futures with Our Nonhuman Symbionts because I am deeply concerned about the widespread assumptions driving our planet into ecological collapse. Chief among them is anthropocentrism: the belief that humans are superior to all other species and are entitled to exploit Earth’s resources for our benefit, regardless of the harm to other species and the environment. If we probe the basis for this belief, it boils down to our superior cognition. We are not the fastest species, or the strongest, or the ones with the most acute senses; but these deficits are more than compensated for, the argument goes, because we are smart enough to invent machines that go faster than cheetah, are more powerful than elephants, and have senses more sensitive then grizzly noses. Combating anthropocentrism, then, requires recontextualizing human cognition in relation to other species and, increasingly, to AIs as well.

The approach I take is to compare and contrast the ways in humans construct our world—our umwelt—with the umwelten of other species and AIs. Central to this enterprise is what I call the Integrated Cognitive Framework (ICF), which locates human cognitive capabilities in relation to those of other biological and synthetic cognizers. In my view, all biological lifeforms have cognitive capabilities, even plants and microorganisms, as do advanced forms of computational media such as Large Language Models. Despite all the risks of our multiple environmental crises and the technological dangers of developing AI, I am committed to envisioning positive possibilities for our human futures. For what is the alternative? To give up and go join a desperate party like those depicted in Poe’s Masque of the Red Death? To be able to enact positive futures, we first have to believe that they are possible.

What is an extract from your book that best represents yourself?

N.K.H.: This passage from the introduction expresses some of the main ideas.

“The Integrated Cognitive Framework (ICF) built on the premise that all lifeforms have ways to sense their environments, absorb information from them, and interpret that information through their sensory and organismic capacities. Thus all lifeforms, including those without central nervous systems such as nematode worms and maidenhair ferns, have cognitive capabilities. Moreover, the interpretations that organisms perform on environmental information are crucial to their survival and reproduction and, in this sense, have meanings relevant to organisms and their milieux. The research area that has most extensively developed this concept of meaning is biosemiotics. Working through the triadic framework of Peircean semiotics consisting of sign vehicle, interpretant, and representamen, biosemiotics understands “meaning” not as an abstract concept but rather as a response to environmental stimuli. As Wendy Wheeler put it, “meaning is always a kind of doing” (2016, p. 7). This conceptual move has enormous implications, for it opens the entire realm of biota to the creation, exchange, and interpretation of signs, and thus to meaning-making practices. Humans in this view are not unique or special in their emphasis on meaning creation; all living beings create, process, and interpret meanings. The uniqueness of humans rather lies in their ability to extend meaning making into abstract symbols, as Terrence W. Deacon argues (1998). Moreover, the signs that living creatures create are performed in environments that include all the other signs. Like the birdsongs and creaturely cries that greet the dawn in an Amazonian rainforest, the signs intermingle, each influencing and being influenced by the others. Jesper Hoffmeyer calls this realm of interacting signs the semiosphere, a grand symphony across scales audible and inaudible produced by creatures as they go about the business of living (1997, vii).

Actors and Agents

The emphasis on cognition has another important implication, for it provides a principled way to distinguish between cognitive acts and material processes. By material processes I mean the physicochemical processes that are the foundations on which all life is built, from chemical reactions in the soil and oceans, on up to the nuclear fission reactions that fuel the Sun. Make no mistake: these processes have agency. Indeed, they often release powerful forces that dwarf anything humans can do. What they do not have is the ability to interpret information and make choices (or selections) based on those interpretations. A tornado cannot choose to plow through a field rather than a populous town; an avalanche will not hold back its pounding cascades because climbers are on the slope. It took the spark of life to create those possibilities, and thus to unleash the power of signs and meaning creation. To encode this distinction, I use the term “agents” to refer to material processes, reserving the term “actors” for living organisms and other entities that embody cognitive capabilities.

The agent/actor distinction is further authorized by the different temporal regimes occupied by material processes and the evolution of biological lifeforms. Physicochemical processes follow trajectories that can be predicted by mapping them into phase spaces; in chapter 5, this is explained by mapping the trajectory of a swinging pendulum into a phase space showing momenta and position. In a seminal article, Giuseppe Longo and Stuart Kauffman (2012) show how this is accomplished even for chaotic nonlinear systems such as a double-jointed pendulum. In contrast, they argue, biological evolution follows no such predictable trajectories. Rather, it leaps from one niche to another in a pattern that Kauffman elsewhere calls “adjacent possibles” (Kauffman 2000, 22).

The unpredictable nature of biological evolution points to another systemic difference between biological lifeforms and physicochemical process: living creatures have stakes in what happens to them, whereas material processes do not, a fact that Michel Levin and Daniel Dennett (2020) emphasize in their subtitle, which refers to “agents with agendas.” Every living being, even a unicellular organism, will act so as to continue its existence; as Darwin noted a century and a half ago, the need to survive and reproduce is universal throughout the animal and plant kingdoms. I call this desire to survive the biological mandate. Through the billions of years of the evolution of life on Earth, all creatures incorporated the biological mandate into their ways of being in world, for if a hypothetical species did not, it would cease to exist and go extinct. Rocks decay, mountains erode, lakes evaporate, but they neither know nor care that these events happen; they simply are.”

What are the trends that are just emerging and that you believe in the most?

N.K.H.: The trend emerging now that I think will have momentous consequences for human evolution is the development of advanced AI such as Large Language Models (LLMs). In my view, AI is the most important cultural adaptation since the invention of language. Granted there are risks involved, but there are also enormous benefits. Computational media have already become our inextricable symbionts. By “computational media” I mean much more than desktop computers and smart phones; it includes the associated infrastructure such as chips, routers, optic cables, server farms, and all the rest of it. If this infrastructure were disabled at midnight tonight, the complex systems on which developed societies absolutely depend would grind to a halt. Banking systems, food distribution networks, cars, trains, planes, the electrical grid, and everything else in which computational media control, process and store information—would all instantly become inoperable. Within six months, the die-off in human communities would be immense. In short, we are already at a point where our complex societies cannot survive without our computational symbionts.

Symbiosis, the joining of two species together, usually implies a benign cooperation, but as a technical term in biology, it also includes parasitism, when one species benefits at another’s expense. In short, symbiosis denotes risks as well as benefits. We are already aware of the risks involved with becoming symbionts with computational media, such as hacking, cyberwarfare, malicious software and so on. It should be no surprise that the risks with AI will be even greater, including the prospect that they will begin to follow their own agendas rather than remain within the goals that humans set for them. However, AI also holds out the promise of vastly increasing the power and reach of cognitive assemblages, the collectivities that include humans, nonhumans and technical entities that are responsible for getting done most of the world’s work. I think that the best way forward is to develop strong international regulatory frameworks that can help keep AI development aligned with human values and human interests.

If you have to give one piece of advice to a reader of this article, what would it be?

N.K.H.: I actually have two pieces of advice. One is to educate yourself with scientifically sound, peer reviewed work about where we actually are now in relation to global warming, environmental damages to the lands and oceans, and extinctions of other species. The second is to acquaint yourself with the LLMs presently available such as ChatGPT, Gemini, Claude and others. Judge for yourself what are the strengths and limitations of these models. Do they make sense to you? Do you find them capable of sound reasoning and meaning creation? When do these capabilities seem to break down? And finally, how can you use them to help achieve the goals that are important to you?

In a nutshell, what are the next topics that you will be passionate about?

N.K.H.: I am already embarked on trying to understand more fully what the advent of LLMs mean for my specific field of literary studies, which I think has not even begun to come to terms with the huge implications of this development for the study and practice of literatures. I am also passionately involved in exploring the extent to which the biological world employs analog computations as it goes about the business of living. Even though digital computation is so pervasive that the “digital” adjective is often omitted (as if all computation were digital), the digital always begins with abstractions that reduce everything to binary symbols and thus is inaccessible to conscious and nonconscious lifeforms that do not perform abstractions. By contrast, analog computation is carried out through continuously changing physical variables interacting directly with each other—a methodology completely within the capacities of conscious and nonconscious organism. What differences would it make if we realized that every living creature around us, including our own brains and bodies, routinely perform analog computations simply to survive?

Since computation is a highly privileged term in our complex societies (for reasons made clear above), a framework that highlights the pervasive nature of analog computation in the biological realm would facilitate projects that combine digital and analog characteristics to reap the benefits of both. It would also increase human respect for the complex ways in which nonhuman creatures sense, process, and interpret information from their environments, thus working hand in glove with the Integrated Cognitive Framework discussed earlier.

Thank you N. Katherine Hayles

Thank you Bertrand Jouvenot

The book: Bacteria to AI, N. Katherine Hayles, First Edition, 2025.