manτεία (manteia)

Roberto Alonso Trillo and Peter Nelson work at the intersection of theoretical research on AI and applied artistic practice. This keynote draws on their experience producing large-scale performances augmented and thematically driven by AI to offer critical insights into AI aesthetics. It examines prior framings of AI art as conceptual (concerned with being about AI rather than of AI - Walshe, 2023) and argues that as AI becomes fully absorbed into the substrate of software and digital practice, such a distinction will cease to carry any critical weight. Into this post-ubiquity landscape, Trillo and Nelson propose that AI aesthetics will shift from speculative fetishisation toward a more productive tension: the contrast between inferential generation and the deterministic algorithmic filtering of more established digital methods. The prevailing fixation on bias and hallucination, they suggest, is better understood in relation to the deterministic processes we have grown accustomed to, rather than as aberrations to be corrected. Moving beyond technical diagnostics, they propose a framework grounded in the concept of apophenia (Bratton, Steyerl) to articulate how artists and audiences synthesise these contrasting computational modes to construct meaning that is both algorithmic and inferential.