Local optima, a concept from the fields of machine learning and neural networks, refers to a situation where the best solution to a problem found based on a limited amount of data turns out to be false. This discrepancy arises not due to an error in the computational method but only becomes apparent after updating the dataset with new information.

The exhibition “Local Optima”, addressing themes of advancing digitalization, transhumanist motifs, and post-internet culture, explores how our visions of the future can prove to be entirely misplaced, despite appearing inevitable. Ideas such as flying cars or teleportation, once at the bleeding edge of 20th-century futurism, in 2024 evoke little more than a wry smile.

Connected by the pulsating network of the internet, we obsessively document everything and have access to almost any piece of information. Into the gaps between existing data, tools like ChatGPT are beginning to pour. However, it is impossible for us to ever predict the future with complete certainty, as our perception is always limited to events that have already occurred or are currently unfolding. Technological advancements, including bioengineering and artificial intelligence, give us the sense that perhaps the moment is near when we will no longer need to guess.

Yet, despite our constant and sincere efforts, it may turn out that even our most probable and rationally supported scenarios of future utopias or catastrophes might just be another local optimum. Until the future arrives, all we can do is document every version of it.

— Wiktor Gałka