N. Katherine Hayles, Bacteria to AI: Futures with our Nonhuman Symbionts (Chicago: University of Chicago Press, 2025) 304 pp. $27.50 Pb. ISBN: 9780226837475
As a species, we Homo sapiens often take it for granted that humanity is separate from, and superior to, the rest of existence. It’s a view that has permeated much of our mainstream thinking over the years, in biology and religion for example, and it still influences the way we tend to think about new issues in biotechnology and AI ethics. In her latest publication, the literary critic N. Katherine Hayles puts forward a holistic alternative to what she frames as misguided and dangerous anthropocentrism, a human-centred outlook that has facilitated the climate crisis and could lead to our ultimate demise. As Hayles sees it, we are just one type of being in a world of cognition that encompasses all living beings and some artificial ones too.
Having previously written on posthumanism (How We Became Posthuman) and non-conscious cognition (Unthought: The Power of the Cognitive Nonconscious), Hayles is on familiar ground in Bacteria to AI: Futures with our Nonhuman Symbionts, and with the exponential rise of AI, the themes she weaves together here are timely. As with the rest of her oeuvre, the ground she covers is eclectic, with time spent analysing views from a wide spectrum of different fields. Notable inclusions are the theories of microbiologist Lynn Margulis; the philosophy of meaning and Peircean semiotics; feminist and eco philosophy; technical explanations of neural nets; literary analysis of works by Ian McEwan and Kazuo Ishiguro; and a critique of writer Yuval Noah Harari. As such, this book will be of interest to a diverse demographic and is likely to widen the horizons of most readers.
The recurring themes throughout the book are cognition and symbiosis: Hayles argues, convincingly, that non-conscious entities we might not normally deem worthy of cognition are, in fact, capable of making and acting on meaning, and that we humans depend on such non-conscious cognition for our continued survival. It’s not simply that we as organisms rely on the biological ecosystems that feed and support us. Rather, in this age of AI and biotechnology, we depend cognitively on cognitive entities – both biological and artificial. Hayles foresees a future where artificial cognisors are even more closely intertwined with our own existence, in which our agency is distributed in a technosymbiotic way, rather than being humanly autonomous.
Our symbiotic relationship with other cognitive entities isn’t just about silicon-based technology, however. For Hayles, biological cognition begins at the cellular level, where microorganisms are capable of deriving meaning from their environments and acting on that meaning. Seen through this lens, scientific breakthroughs such as vaccines and gene editing are less a feat of human endeavour and more a collaborative effort combining both human and bacterial cognition – which neither agent could achieve alone. To understand this way of thinking, we need to step through the looking glass, inverting our long-held assumptions about our place in the world. One way to do this is to apply ‘reversible internality’ to see that the anthropocentric view is just one way of looking at things.
In Chapter 4 for example, Hayles invites us to apply reversible internality to the practice of CRISPR-Cas9 gene editing. This ingenious technique harnesses the natural mechanisms by which bacteria build immunity to viruses and has been used by scientists in treatments for sickle cell anaemia and some cancers. In nature, when a virus threatens bacteria, the bacteria copies and incorporates the virus’s DNA, subsequently using these same DNA sequences to attack and kill the virus. This, Hayles argues, demonstrates the abilities of even microscopic organisms to derive meaning from their environment and use it to direct their actions. The CRISPR-Cas9 gene editing process uses the inherent cognitive capacity of bacteria, enabling scientists to make repairs to disease-causing gene mutations. While it’s tempting to think of human scientists as the only heroes here, we can also change our perspective to recognise the indispensable role of bacterial nonconscious cognition in such technology.
Hayles doesn’t address artificial intelligence in depth until the later chapters, using the first section of the book to focus on living organisms. In Chapter 6, she finally turns her focus to how large language models like GPT-3 (the technology behind ChatGPT) can make meaning. In such a fast-moving field, it’s inevitable that newer technologies will quickly succeed those described in print, and some elements of this discussion – such as the exploration of GPT-3’s literary capabilities – already feel slightly dated. Nevertheless, it’s interesting to see Hayles apply her expertise and perspective as a literary critic to the outputs of an LLM, and her thoughts on why this is valuable.
Later, in Chapter 7, the discussion moves to GPT-4 and an exploration of what it means for AI to be creative. We humans tend to look at this from a parochial perspective and either dismiss the possibility of AI creativity altogether, or judge it on our own terms. But what if AI were creative within its own remit – constructing ‘new interesting problems they had identified in the arena of data compression and analysis?’ (175) she muses: ‘Would the area of AI “indigenous” art be something that only AIs could appreciate and evaluate?’ (175). To truly reap the benefits of AI’s creative potential, Hayles advocates a collaborative attitude, allowing AI systems to investigate the areas they find interesting and reject those they don’t. After all, truly creative inspiration doesn’t emerge from rule-following and compliance. She develops this idea throughout the remaining chapters, exploring what robot consciousness might be like, and emphasizing that both humans and artificial cognisors have their own considerable, compatible strengths. The book ends by addressing the theme of reversal yet again, putting forward three ideas for changing our current, human-centred course.
With a subversive message that turns traditional thinking on its head, Bacteria to AI: Futures with our Nonhuman Symbionts outlines a surprisingly compelling new perspective on the role of humans in the world. Penned by a polymath with expertise in both chemistry and literary criticism, its eclectic content perhaps reflects the unconventional, integrated approach that the author advocates for our future existence, breaking down disciplinary boundaries and synthesising latent meaning from truths we have known all along. It certainly prompts the reader to re-evaluate some deeply-held conventions, offering a provocative challenge to human-centred thinking.
Catherine Bowden, Northeastern University London
