There seems little doubt that the integration of artificial intelligence (AI) into computational architecture is transforming how students learn, develop, and evaluate design. AI can accelerate coding, modelling, visualisation, optimisation, enabling students to explore possibilities that might otherwise require considerable time and technical effort. Yet this acceleration raises questions about what should be automated, what should remain cognitively demanding, and what level of resolution is necessary for learning and design decision-making. I suggest that effective use of AI is not simply about maximising automation, speed, or output, but about deciding when to accelerate, what to automate, and when to slow down. High-resolution may be unnecessary during early conceptual exploration, where rapid, low-resolution can encourage experimentation and reflection. Similarly, automated code generation can increase speed while reducing opportunities to understand algorithms, debug processes, and reason through computational relationships. This also prompts a reconsideration of what is meant by computational. Computational does not necessarily mean digital, just as digital does not necessarily mean computational. Some architectural ideas are intrinsically digital, while computational thinking can also be embedded in analogue drawing and physical modelling. Precedents demonstrate that complex geometric and spatial outcomes can be achieved through such means. Within this context, I ask whether hand-sketching still has a role in architectural education. My suggestion is that its slower, imprecise, and embodied qualities can support ambiguity, spatial reasoning, and reflective decision-making. I conclude by proposing a pedagogy of purposeful acceleration, variable resolution, and deliberate friction, in which AI and analogue practices become complementary means of developing critical, computational, and architectural judgement.
Nerma Cridge is an architect and academic whose practice operates across architecture, art and spatial design. Educated initially in Sarajevo, she continued her studies in Birmingham and The Bartlett before completing her PhD, Drawing the Unbuildable, at the Architectural Association in London, where she continues to teach. Through her practice – Drawing Agency – she explores drawing as a generative and performative tool, combining computational, material and experimental approaches to produce temporary spatial conditions and different forms of architectural representation.