The three fates of AI: Convergent, divergent, deviant
The same AI can simultaneously stimulate and diminish human interaction and coordination, depending on how its possibilities are used.
Artificial intelligence (AI) tools do not have a single fate when distributed among employees; instead, they can converge, diverge, or deviate. Convergence occurs when a tool's use is standardized among a thousand employees, leading to tighter coordination. Divergence takes place when different teams use the same tool for various legitimate purposes, some visible and others hidden. Deviance involves using the tool in ways that deviate from an organization's policies, but are still beneficial and goal-oriented.
Convergent AI affordances are engineered, not natural, where someone intervenes to prevent a hundred assistants from blooming and concentrates use onto productionised, governed capability. This leads to increased coordination, as demonstrated by human-in-the-loop philosophies in organizations. Divergent AI affordances, on the other hand, are valuable and can lead to innovation.
Citizen-developer platforms show the pattern at scale, with the same affordance used in thousands of different ways by businesspeople to solve their own problems.
Deviance can manifest in two forms. The classic form arises when tools spread into capable hands, leading to the inappropriate use of both sanctioned and unsanctioned uses. The second form is stranger and more important for governance, involving vendors shipping AI features enabled by default, creating unsanctioned use before any user decides to deviate. Both forms highlight the need for organizations to adapt to changing conditions and manage the fates of AI affordances intentionally.
Written by urgent.news from ITWeb's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
