Research surrounding AI control mainly focuses on non-AI functions with control over AI functions being rare. Available works use AI control to select AI models for a user-specified AI system or to increase the accuracy of an AI system. When AI control is applied in such a fashion, it can increase energy or inference cost (Shen et al., 2023; Wu et al., 2023; Maini et al., 2022). Our article lays foundation for AI-on-AI control that reduces these costs for systems with multiple computer vision functions. SICEC evaluates an input and only activates the relevant system functions for that input. SICEC also attempts to gauge the complexity of an image and assign lower cost function-related models when possible. Results promote the viability of cost-reductive AI-on-AI control research showing significant energy and inference time reductions. SICEC like methodology could be applied to increase the long-term sustainability of various AI systems, examples being computer vision cloud, generative...

As artificial intelligence (AI) systems become more complex and widespread, they require significant computational power, increasing energy consumption. Addressing this challenge is essential for ensuring the long-term sustainability of AI technology. AI-on-AI control refers to a system with a set of AI functions controlled by an upper-level AI model. Previous work in AI-on-AI control focuses on boosting accuracy or expanding system capability by increasing overall system cost. Alternatively, we focus on applying AI-on-AI control to decrease system cost and increase the sustainability and viability of a system with multiple AI functions. Our supervised image classification evaluative controller (SICEC) is a cost-reduction oriented AI-on-AI controller that learns when vision models within an AI system should be activated based on input features. The function controller (FC) preprocesses an input and activates relevant functions, functions being distinct units of AI functionality within ...