Information is, among other things, a quantification of the difference between the current state of a system and its most probable state. Artificial Intelligence as proxy precognitive toolkit acquires utility and cultural – if not institutional – value as a function of the extent to which its models approximate to and validate the contingent, transient (systems of belief as) realities that percolate to ascendance in any particular time and place. An enigma here is that the approximation to future states is always and already encapsulated by normative metrics that are themselves the conceptual ancestors of the past and all its biases and inequities. We and our technologies move forwards by also recursively moving backwards in time.
I would be interested to see someone produce a cybernetic interpretation of the movie Tenet. Abstracted above and beyond the particular narratives as teleologies in cultural artefacts, experiences or instances of systems that embody a transmission medium for our shared historical moment, the overall arc and trajectory of algorithmic pattern generation looks, Janus-like, to both past and future. We weave a future as we unravel the past but the past we choose to see shapes the future we expect. Our errors are axiomatic.