This month’s Frame: using rhizomatic thinking to unlock new ideas
A framework to nurture new creative frontiers using Generative AI.
Generative AI (Gen AI), a subset of artificial intelligence, focuses on creating new content, such as text, images, and music, based on patterns in data. The dominant narrative around Gen AI often focuses on how well it can replicate human abilities. From chatbots designed to sound convincingly human to image generators that aim for photorealistic precision, much of the discourse is centred on imitation.
The fascination with mimicry is deeply embedded in our cultural imagination, as seen in iconic representations of AI in films like Blade Runner and Ex Machina. Such portrayals reinforce the idea that AI’s primary value lies in how closely it mirrors human abilities. Moreover, the emphasis on mimicry amplifies societal anxieties about AI’s role in creating widespread unemployment, as it feeds fears of machines not just replacing human tasks but also encroaching on what makes us distinctly human.
When Gen AI tools deviate from expected outputs—producing so-called “hallucinations” or aberrations—these are often dismissed as mistakes or flaws to be corrected. Viewing these deviations as errors is short-sighted. Such an outlook reduces AI’s potential to mere imitation, overlooking its ability to create something entirely new.
Instead of focusing on replication, we should consider how Gen AI can leverage its unique capacities to make unexpected connections, uncovering possibilities that humans might not imagine on their own. A change in perspective could open up new frontiers for creativity and innovation.
The framework
We can apply Deleuze and Guattari’s concept of arborescent and rhizomatic thinking to reimagine how we use Gen AI. They contrast two approaches:
Arborescent thinking: Represented by a tree with a central root anchoring everything, this approach is hierarchical, linear, and focused on following predefined paths toward a single, ideal outcome. Such thinking prioritises replication, order, and stability. The current discourse around AI mirrors the arborescent approach, fixating on predictable outcomes, the mimicry of human abilities, and the refinement of existing systems to achieve predictable outcomes.
Rhizomatic thinking: This is like a rhizome, a root system that grows unpredictably in all directions, forming connections without a central root or hierarchy. Rhizomatic thinking is decentralised, fluid, and focused on exploration. It values multiplicity and embraces the unexpected, allowing for creativity and the discovery of entirely new possibilities. Rhizomatic thinking offers an alternative way to think about what we can achieve with AI.
In their critique, Deleuze and Guattari argue that arborescent thinking stifles creativity by restricting it to predefined structures, hierarchies, and existing patterns. As they famously put it, “we hate trees,” rejecting the rigidity of linear, rooted systems that prioritise order over exploration. In contrast, rhizomatic thinking invites us to embrace uncertainty, divergence, and complexity—qualities that foster groundbreaking ideas and unexpected connections. Rather than focusing on what something represents or imitates, a rhizomatic perspective challenges us to ask a more open-ended question: “What can it do?”. It encourages a mindset of experimentation and discovery, unlocking possibilities beyond the confines of tradition.
Applying Deleuze and Guattari’s rhizomatic thinking to Gen AI allows us to rethink how we design, use, and understand these systems. Instead of focusing on AI as a tool for predictable, predefined outcomes, their framework encourages us to embrace its potential for making unlikely connections and generating surprising, even chaotic, results.
A powerful analogy for this approach finds itself in the work of abstract artists like Jackson Pollock, who embraced unpredictability as a central element of their creative process. Pollock didn’t paint to faithfully replicate reality but instead explored the capabilities of his medium—letting randomness, movement, and spontaneity guide the emergence of entirely new artistic forms.
Similarly, AI need not strive to mimic human creativity or understanding but can instead chart its own path, generating novel possibilities that both inspire and provoke us. By embracing the unexpected, AI becomes not just a tool, but a collaborator in pushing the boundaries of imagination and discovery.
Using the framework
Across many different types of work, from the most obscure art practices to the most technical engineering workflows, there is usually a “generation” phase. This requires high volume, low-fidelity output—a fashion designer might create an entire book of sketches, a copywriter a raft of taglines, a researcher a list of hypotheses to prove / disprove. During this “front end” phase, there is often value to be found in exploring beyond the initially obvious, making forays into unexplored territory.
Within an innovation process, what if instead of using Gen AI to automate repetitive tasks or imitate human creativity, we could lean into its potential to expand this sort of “front end” work? This is where large datasets, when processed by AI, could offer unparalleled opportunities to uncover intricate patterns, subtle relationships, and unexpected connections that would otherwise remain hidden to humans.
In other words, humans simply can't retain such volumes of information in their minds to form the kind of cross-pollinating connections that AI can achieve, enabling discoveries (or at the very least, initial concepts / thought starters) far beyond the reach of human intuition or traditional analysis. That initial generative, low-fidelity exploration could then be synthesised and deepened by skilled human workers.
A notable example of this kind of approach in action is Philippe Starck’s collaboration with Autodesk, where AI was involved with the design of a new chair, using minimal material and energy. Instead of delivering conventional results, the AI explored millions of possibilities, generating lightweight, organic forms that humans might never have considered. It didn’t replace the human role in design but enhanced it, providing fresh ideas that were later refined and made practical.
Instead of seeing AI “hallucinations” as flaws to be corrected, we could treat them as features—sources of inspiration and new ways of thinking. By letting go of rigid expectations and embracing unpredictability, we unlock AI’s rhizomatic potential to create entirely new possibilities rather than simply perfecting the familiar.
NEWS
We recently had the pleasure of speaking to Professor Neil Lawrence on the publication of his book, “The Atomic Human: Understanding Ourselves in the Age of AI”. We touched on historical precedents for big technologically-driven shifts, his view on how to develop AI to be most beneficial to society and what professions need to focus on to survive the coming changes. Read the full interview.
Frames is a monthly newsletter that sheds light on the most important issues concerning business and technology.
Stripe Partners helps technology-led businesses invent better futures. To discuss how we can help you, get in touch.
