meaning-making
I’ve been stuck for the last few days on the gap between information and meaning, spurred by this fascinating essay Advait Arun. He analyzes the economic qualities of information as a foundation for thinking through claims about the impact of AI. Most critically, he argues, better systems for information processing do not close the critical gap to acting on information:
This could very well be the existential condition of information processing: No information system, as advanced as it may seem, will completely bridge the fundamental, last-mile incommensurability between how information is disseminated and how it is received.
New technologies may not, at any point, surmount the cognitive challenges of identifying valuable information, imbuing it with meaning, or acting on it.”
I’ve been reflecting on this in the context of my work. I think there are two fundamental skills embedded in the type of analytical work I do: 1) producing valid information and 2) making meaning from valid information.
The first skill is process oriented - chaining together a series of logically coherent steps that transform a series of solid inputs using reasonable assumptions into a clear output signal.
The second skill is relational and steeped in context. It requires knowing your audience deeply, understanding their desires and incentives, their mental models of the world. The core skill is situating information within their understanding in a way that helps them take concrete action based on the new input.
I think the first skill is going to be fundamentally transformed by AI, and is already changing what we need from more junior analysts. But meaning-making seems much more likely to stay stubbornly human, particularly when it requires trust and mutuality.