Proxies - Intuition Stories - Speculative Machines and Us

intuition
stories

Stories for teaching, learning, and research: From electric brains to generative AI, explore intuition in digital computing histories through 40 key concepts, initiatives, milestones, systems, and publications. Uncover how questions about sensing, thinking, and speculating animate current debates on machine intelligence.
Proxies
An entity that stands in for another entity; common in modelling and machine learning
Proxies

Proxies

Utilised in different kinds of modelling, a proxy is an entity that takes the place of another entity, or is understood to have the same meaning or purpose as it.

When something cannot be modelled or calculated directly, a proxy may be used to approximate or “intuit” its dynamics amid uncertainty. As the media scholar Dylan Mulvin suggests, proxies bear ‘the promise of creating controllable renditions of an unpredictable and unknowable’ (2021: 4).

Weather and climate models often use proxies. Paleoclimatologists, for instance, mobilise proxies for temperature ‘such as tree ring measurements, ice cores, ice melts, and local weather records’ (Chun, 2021: 127); a necessary practice in the face of incomplete or uncertain data which has informed the development of Earth Systems Models (Lynch, 2008; Edwards, 2010).

Within machine learning-enabled “artificial intuition”, proxies expand ‘the estimable beyond the calculable’ (Hoyng, 2025: 9) for the purpose of discovery. In this context, and more generally, proxies both ‘reduce and introduce uncertainty’: they enable researchers to expand what is knowable yet, in doing so, introduce new forms of indeterminacy (Chun, 2021: 125).

As a ‘culturally conditioned practice of consistently using some things to stand in for the world’ (Mulvin, 2021: 5), “proxification”, however, can also reinforce stereotypes and inequalities.

As the digital media scholar Wendy Chun suggests, statistical techniques employed within machine learning systems to “intuit” previously unknown associations, such as K-means testing, for example, draw on proxies ‘to compensate for ignorance or lack of evidence’ (2021: 57). Proxies used in social data analytics for marketing, policing, insurance provision, and electoral engineering (e.g. the use of postal code or ‘age of first arrest’ to infer race or brand preferences to infer ‘low IQ’) are often discriminatory and entangled with eugenic histories.

The fact that proxies are vital to understanding urgent phenomena such as climate change but also underlie discriminatory analytics underscores their fundamental ambivalence – and that of correlative modes of intuiting, knowing, and predicting more widely.

Artificial intuition
AI concept proposing how machines might simulate intuition for rapid, creative, or uncertain decisions
Artificial intuition
Neural network
AI architecture inspired by brains, enabling learning from data through layered connections
Neural network
Post-cybernetic logic
Leveraging computational uncertainty to generate value
Post-cybernetic logic