What makes a paradigm the common ground for expert insiders?
This explores why a shared paradigm is something expert communities hold in common — what social machinery makes it the agreed ground insiders stand on, and why that's something AI can read about but not actually join.
This explores why a shared paradigm is something expert communities hold in common, and the corpus has a clear, slightly surprising answer: a paradigm isn't common ground because everyone learned the same facts — it's common ground because it's validated through participation, reputation, and the ongoing negotiation of who counts as an insider. Expertise, the collection argues, is socially validated through community membership and track record, not individual accuracy Can AI ever gain expert community trust through participation?. The paradigm is the consensus-building process itself, which is why an outsider — or an AI — can master every claim inside it and still not belong to it.
The interesting move is that the force of an idea inside a paradigm comes partly from who is saying it. Arguments carry weight not only on their content but on the standing of the thinker — reputation, history, position in the field Can language models distinguish expert arguments from common assumptions?. That's why language models, which process text but not the social world where authority is built, struggle to tell an expert claim from a widely repeated assumption. The common ground isn't just shared propositions; it's a shared map of whose word counts on what.
And holding that ground is active work, not passive agreement. Communicative grounding shows that even shared words don't guarantee shared reference — insiders have to keep calibrating what terms mean to each other Why do speakers need to actively calibrate shared reference?. Expertise turns out to be communicative at its core: an expert judgment always anticipates its audience and what that audience will accept as valid Can AI replicate the communicative work experts do?. The paradigm is less a fixed body of knowledge than a continuously maintained agreement about acceptability.
That reframes what expertise even is. Several notes converge on the idea that being an expert is a role performance — knowing when to speak, when to defer, which knowledge applies right now — rather than just possessing a bigger stock of facts Is expertise really just knowing more than others?. This is also where the cognitive-diversity finding bites: diverse teams only produce better ideas when members already share foundational domain expertise; without that common ground, diversity becomes noise that degrades the work Does cognitive diversity alone improve multi-agent ideation quality?. The paradigm is the precondition that makes productive disagreement possible.
The doorway worth walking through: this is exactly the boundary AI can't cross. Because the paradigm is constituted by social embeddedness and a testable history of judgment, a system that only emits fluent text is structurally locked out of the validation circle, however accurate it sounds Does AI separate intellectual form from the thinking behind it?. The common ground of insiders isn't information — it's membership, and membership is something you participate your way into, not something you retrieve.
Sources 7 notes
Expertise is validated through social participation and track record within expert communities, not individual accuracy alone. AI cannot enter this validation circle because it lacks social embeddedness, testable judgment history, and ability to participate in the consensus-building processes that define expert paradigms.
LLMs lose the social context that gives expert claims their force—reputation, track record, and standing—because they process only text, not the social world where expertise is built and evaluated.
The same words can mean different things to different speakers because referential grounding is person-specific. True communicative grounding demands collaborative negotiation of how language connects to the world, not mere surface-level word sharing.
Expertise requires anticipating audience acceptability and social validity, not just retrieving information. AI lacks the mechanism to perform this communicative work, making its fluent output epistemically misleading despite its confident form.
Real expertise involves situational judgment—knowing when to speak, when to defer, which knowledge applies now, and how to communicate it to a specific audience. This role-performance dimension is at least as important as the underlying knowledge stock, and it is what AI cannot structurally perform.
Multi-agent teams substantially outperform solo ideation, but only when members possess genuine senior knowledge. Diverse teams without expertise underperform even a single competent agent, because cognitive stimulation without expertise triggers process losses instead of insight.
Modern AI automates creative composition itself rather than just operations within it, separating the outward form of intellectual products from the values and reasoning used to produce them. This mechanism allows exchange value to float free from use value.