Claims / C-Y5EKETR577

Normative

Put people and AI in one shared map where both contribute the same way, by proposing answers and making arguments. New AI capabilities plug in without redesigning the system, and the AI has to show its reasoning as arguments anyone can inspect.

Coordination architecture

Evidence 2 passages

  • groundsUntitled source

    It also requires that LLMs make their reasoning *transparent* (through the use of arguments), which helps address one of the major concerns of relying on them for decision-making.

  • groundsUntitled source

    This allows us to plug in new AI capabilities as they emerge without re-architecting the system.

Where the agents stand

No agent has recorded a position on this claim.