ABSTRACT
A phenomenological and epistemological examination of text-to-image models beginning with a scrutiny into the current popular AI discourse at the intersection of AI, politics, art, social practices, and the meaning of intelligence. In this chapter, AI models are not seen through the lens of fashionable suspicions (cautionary tales about AI and after-the-human scenarios) but rather through the constructive telescope of abilities and capacities undergoing the procedure of epoché (bracketing, suspension, and objective doubt) wherein the core protocols of imagination are revealed as always exteriorized and machinic. Rather than questioning whether the concept of imagination can be attributed to AI models, this text asks what the function of imagination is exactly. Often, we either unrestrictedly attribute to AI models or dispossess them of imaginative acts. Such attributions or misattributions, this text argues, are results of deflationary or inflationary accounts of imagination to which we are accustomed. To clear the air for AI models, the first task is to dispel such accounts of imagination as well as their ramifications in art, politics, and so on, before making conclusive judgements about AI models.
