Note on Images

This website features original photographs as well as visual images and videos produced using generative AI models with prompts that explore AI concepts and histories.

The use of AI-generated images raises complex epistemological, political, and ethical issues in our current digital media landscape. These include, but are not limited to, concerns regarding copyright, intellectual property, and plagiarism; labour and capitalist extraction; “deep fakes”, truth, and accuracy; trust and the integrity of knowledge eco-systems; privacy, vulnerability, and abuse; algorithmic bias and injustice; and the politics of representation.

The process of producing visual outputs using generative AI, also, however, illuminates issues at the heart of the Speculative Machines and Us research project, such as those related to the changing dynamics of human-algorithm interaction; computational common sense; machine learning “hunches”; distributed sensing, cognition and intuition; and algorithmic structures of feeling.

To generate a varied collection of images and videos using Large Language Models and Generative Image Models, we engaged in an iterative process of prompt engineering and refinement that attended to priorities of: visual interest and engagement; historical and geo-political context; plausibility and integrity; social inclusion and diversity; and socio-technical creativity and imagination. These priorities also informed our final selection of images from a much larger collection of visual outputs.

The guiding idea was to explore these new generative AI models openly, in order to see how much of the creative process could be managed by AI systems. The intent was to test the degree that anyone could now use a detailed prompt to create images “stylistically” themed to match the overall “brand” of a website.

After significant experimentation, this prompt was used to interrogate the most popular Large Language Models (such as ChatGPT, Claude, Perplexity, Grok, DeepSeek, and Mistral) for creative interpretations of the topics and themes featured on the website, and particularly in the Intuition Stories section. The resulting prompts were then explored within a wide range of Generative Image Models (such as Flux, Nano Banana, Ideogram, Lucid Origin, GPT‑Image, Grok, and Krea). These models were also used to complete retouching processes traditionally handled in applications like Photoshop, such as cropping and enlarging images and adapting content. Finally, some of the stronger results were further developed using Generative Video Models such as Sora.

Such creative and computational processes, we suggest, point to the asymmetric entanglement of human and algorithmic tendencies, sensibilities, judgements, and decision-making capacities in the age of generative AI.

We invite visitors to approach the images and videos featured on this site as provocations that may spark useful questions, critiques, and debates concerning the possibilities, limits, and risks of generative AI, the logics of algorithmic culture, and the workings of “artificial intuition”.

Carolyn Pedwell and Jason Rainbird (Rainbird Digital)