Idea Ripper // ArchiveCards: 1555 [Live] | Updated: 2026-10-14
Card #c14IR // #c14Synthesis

The technically and institutionally mature objective is bounded autonomy : the agent can plan, retrieve, calculate, and propose actions inside a formally defined action set, while external shields enforce limits and humans retain authority over material decisions.

— In our words · Source: Large Language Models in Finance (Miquel Noguer i Alonso)

[Use when]

Use when deciding what an AI agent is permitted to do in production: which actions stay read-only, which need review gates, which are off-limits.

[Why it matters]

Fluency is the trait everyone watches improve, but it says nothing about whether an agent should be allowed to move money. The book's growth rule: the action set expands only after evidence of reliability, never because the model got more fluent — permissions follow demonstrated trustworthiness, not convincing prose.

Related principles

#c13Prediction and agency should not be conflated. A credit model can estimate the probability that a borrower breaches a covenant. An agentic…Use when a team proposes a stronger model as the fix for an agent that lacks memory, permissioning, or an…

More from Large Language Models in Finance

#c13Prediction and agency should not be conflated. A credit model can estimate the probability that a borrower breaches a covenant. An agentic…Use when a team proposes a stronger model as the fix for an agent that lacks memory, permissioning, or an… #c15This principle prevents a common category error. A model that says it will respect a mandate has not yet respected the mandate. A model…Use when a vendor demo, benchmark score, or red-team chat log is being offered as proof that an agent is safe… #c16Production systems should separate language generation from numerical computation. Let the LLM propose hypotheses, retrieve context, or…Use when an LLM-generated forecast, ratio, or expected return is about to enter a report, model, or trade.