A reference companion for the course — the language we'll be using across the eight thinkers, with brief definitions. Use it however serves you: read it through once, dip in mid-session, or come back when a term resurfaces.
Nothing matches that search. Try a shorter word, or clear the filter.
The stance we're working from, not the thinking of any single thinker.
A place to think about AI that isn't captured by either polarised camp. Neither breathless optimism nor existential dread. The whole course is an attempt to open this space and learn to stand in it, aligned with our mission at 3rd Space.
The practice the course is built around: refusing the rush to certainty (in either direction), held in a discussion-based setting — a place where we can ask questions without having the answers ready to hand.
Biologist whose theory of symbiogenesis reframed evolution. Cooperation, not just competition, drives the greatest leaps.
The theory that new life forms arise through the merger of existing organisms. The mitochondria in every one of your cells were once free-living bacteria. You are already, biologically, the product of a merger.
Margulis's expanded framework: complex (eukaryotic) cells emerged through a sequence of merger events, each adding a new component. Rejected by mainstream biology for years before becoming foundational.
Margulis's challenge to the strict "survival of the fittest" framing. Competition is real, but the most significant evolutionary leaps happened through cooperation and merger — not through competition winning out, which is more a mechanism of status perhaps.
Not a mystical earth-goddess but a rigorous biological observation, developed with James Lovelock: life and Earth's chemistry have co-evolved into a self-regulating system. Margulis was keen on reclaiming Gaia from spiritualised misreadings.
Writer and theorist whose holon concept reframed how we think about parts, wholes, and the nested architecture of systems.
Every entity is simultaneously a whole unto itself and a part of a larger whole. A cell is a holon. So is an organ, a person, an institution, a forest. There is no top of the hierarchy, no bottom — wholes-within-wholes all the way through.
A system of nested holons. Unlike a hierarchy, a holarchy has no privileged level — every level is both whole and part. Useful for thinking about AI not as something separate but as a new holon emerging in a system we're already part of.
Every holon has two faces, like the Roman god Janus: it asserts itself as a whole, and it integrates as a part of something larger. Self-assertion + integration. These tendencies are always in productive tension — and visible everywhere from cells to societies.
The two opposing pulls in every holon. Too much self-assertion = isolation, fragmentation. Too much integration = loss of identity, conformity. The art is in the balance — and it's a useful lens for thinking about our own relationship to AI.
French philosopher whose work on technology and individuation was largely ignored in his lifetime and is now, we believe, foundational for thinking about our relationship to AI.
Humans and our technologies don't just coexist — they co-constitute each other. The tool shapes the maker as much as the maker shapes the tool. 'Us' and AI are already shaping each other.
Becoming is never finished. To be a self is to be in process — always individuating, never finally individuated. Simondon refuses to start his thinking from finished individuals; he starts from the field of becoming, with individuation as ongoing event.
Technical objects evolve toward greater internal coherence. A mature engine isn't "more complicated" than a prototype — its parts are more necessary to one another. The lens turns "technological progress" into something more rigorous than mere complication.
What exists before individuation: a charged field of potentials, neither subject nor object yet. Simondon's wager is that most modern thought assumes finished individuals as starting points, when in fact we're always still emerging out of the pre-individual.
The space in any technical object where things are not fully determined — where it can respond, adapt, take cues from its environment. Critical for understanding why AI feels different from older technologies: its margin of indetermination is much larger.
Swiss-German philosopher whose framework on the mutation of consciousness is one of the most rigorous tools we have for naming what's happening in our moment.
Gebser identifies five: archaic, magic, mythic, mental-rational, and an emerging integral. They are not stages you grow through and leave behind — they are structures always present and available in us. Each has its own gifts.
Consciousness doesn't develop linearly. It mutates — discontinuously, in leaps. New structures emerge alongside old ones, never replacing them. The distinction matters because it stops us from treating "integral" as a higher class to graduate into, a common place to rest for many of us, particularly in the West.
The dominant structure of modernity: linear, analytical, separative, perspectival. It gives us science, technology, the capacity for control. Necessary. Also, in its current form, in deep crisis.
When the mental-rational operates without integration — pure abstraction, pure separation, pure control. Gebser's diagnosis of the late-modern crisis. You can see it in dominant AI discourse (both utopian and apocalyptic): same structure, different sign.
The emerging structure. Not a higher stage but the integration of all previous structures — archaic, magic, mythic, mental-rational, all alive and available. Capable of holding contradiction without collapsing into one perspective.
Without a single perspective — multiple perspectives held simultaneously without flattening them into one. A capacity of integral consciousness. Distinct from "no perspective" (anything goes) and from "many perspectives juxtaposed" (relativism).
Translucent, see-through. For Gebser, integral consciousness lets the origin show through — not opaque (like the deficient mental-rational), not without form, but luminously transparent. A quality, not a doctrine.
Clock-time is the mental-rational's flattening of time into measurable, identical units. Time-as-experience is time lived, dilated, slowed, accelerated by attention. The same hour can be very different times. Gebser's recovery of this distinction is the spine that connects him back to Simondon. See also: associated milieu, mental-rational structure.
Gebser's title and central idea: the origin isn't behind us in the past — it's with us, always available. Every structure of consciousness is rooted in this ever-present origin. Time, properly understood, isn't linear distance from the origin; it's the origin's ongoing presence.
Jesuit priest and palaeontologist whose lyrical vision anticipated the internet, AI networks, and the conscious Earth decades before they existed.
The "sphere of mind" — a layer of human thought enveloping the Earth, just as the biosphere is the layer of life. Teilhard coined this in the early 20th century. The internet looks uncannily like one possible realisation of it.
Teilhard's proposal: as systems become more complex, consciousness emerges from them. Not a metaphysical claim but a structural correlation between complexity and inner experience. Controversial — and generative for the AI question.
Teilhard's geological layering of Earth: rock, life, mind. Each layer emerges from and is constituted by the previous. Reading him now, with AI networks added, makes you wonder whether another layer is forming.
Teilhard's controversial vision of a final convergence of consciousness. Use carefully — it's the most teleological and religious part of his thought, and the easiest to misread as Silicon Valley singularity. The course holds it as a question, not a prediction.
The process by which the human emerges. For Teilhard, this is ongoing — we haven't finished becoming human. A useful frame for thinking about what AI is doing to that process.
Contemporary philosopher whose argument frames AI as a rupture in the categories we've used for centuries to understand ourselves.
AI doesn't just challenge particular ideas — it breaks the categories we've used to understand ourselves and the world. Human vs. machine is the most obvious one to go. Rees's claim: this is a rupture comparable to historical axial-age shifts.
Rees's analogy to historical axial ages — periods of religious or cosmological transformation across multiple civilisations roughly simultaneously. We are, in his reading, in one. The course holds this as a generative provocation rather than a settled claim.
Capabilities of AI systems that no one designed in. They simply emerge from scale and training. A defining technical feature of contemporary AI — and a source of both wonder and danger.
Brazilian-Canadian scholar whose decolonial and meta-relational work offers a profoundly different way of being with intelligence — including artificial intelligence.
Relating beyond the categories of self/other, human/non-human. Not a position about AI but a posture toward it — and toward everything else. The relation is what's real; the apparent things in relation are partial appearances.
Andreotti's framing: relate to AI as kin — not as tool, not as threat, but as relative. A re-positioning that changes everything about how the relationship can unfold. Uncomfortable on purpose; it asks more of us than mastery or fear do.
The practice of holding modernity — its institutions, certainties, identities — as it dies, without rushing to replace it. Andreotti's broader framework; the AI work emerges from it.
Andreotti's practical method: engage each interaction with AI as a relational act, not a transactional one. Notice what you bring, what arises, what shifts. A productive tension with Simondon's associated milieu, which sees the relation as already constituted. Compare with associated milieu.
Senior figure inside the AI industry and one of the more reflective voices working from within the field. Useful precisely because he doesn't share the course's consciousness perspective.
Every significant technology in history has exceeded what its creators intended. Fire, agriculture, the printing press, the internet. AI just makes this visible faster. The long arc, from someone inside the industry.
The belief that we ever fully controlled our technologies. We didn't. AI just makes that loss of control obvious — but the loss was always there.
Agüera y Arcas's question more than his answer. He doesn't fully commit to whether AI is or isn't intelligent — but he insists the categories we've been using don't fit. The question is generative; the answer is open.
Living with Uncertainty · 3rd Space · 2026. Co-facilitated by Faheem Nusrat & Steve Brett. A living glossary — entries will be refined and added as the course unfolds.
