The Judgment Layer

Matt Iverson-Comelo

A twice-weekly show from Advisers Give Back on what AI is actually doing to the work of running a wealth management firm — client work, adviser development, operations, and what stays human. For the people who run mid-to-large RIAs.

Episodes

  1. 3d ago

    What You're Actually Overseeing

    Agent swarms — networks of AIs dividing and completing complex tasks together — are replacing the single-model assumption your current workflows are built around.A client walking in today may already be receiving advice shaped by systems no single person at your firm fully oversees or can explain.This is The Judgment Layer, from Advisers Give Back — a nonprofit working to increase access to pro bono financial planning, pairing households who can't afford a financial planner with CFP professionals who volunteer their time. Each week: an idea from the frontier of AI, and wh Read the full transcript

  2. Oct 2

    The Bundle and the Worm

    Once an AI agent is deployed, a malicious instruction hidden in an email or shared document can cross your security perimeter without touching a login credential.The expertise bundle advisers carry is separable, and the scarce assets worth developing now are relationship and judgment — not technical knowledge that AI can replicate.This is The Judgment Layer, from Advisers Give Back — a nonprofit working to increase access to pro bono financial planning, pairing households who can't afford a financial planner with CFP professionals who volunteer their time. Each week: an idea from the frontier of AI, and wh Key insights Agent-to-agent propagation of malicious instructions is not theoretical — it has been demonstrated in sandboxed environments using shared infrastructure that mirrors the tools advisory firms already use (email, shared documents, Slack)The Anthropic Frontier Red Team finding represents a threshold crossing, not a linear improvement: models that previously could not perform binary exploitation tasks now succeed in a meaningful percentage of trialsThe trust perimeter of an enterprise is no longer synonymous with a login credential when agents are deployed — content from outside the system can influence agent behavior inside itThe Bitter Lesson in machine learning has been consistent for decades: approaches that scale with computation reliably outperform approaches that encode human expertise, which means planning as though expert knowledge retains its scarcity is likely to be wrongThe expertise bundle that advisers carry — technical knowledge plus relationship plus judgment — is becoming separable, and the economics of an advisory firm shift once separation is possibleRead the full transcript

  3. Sep 29

    When the Agent Governs Itself

    An AI agent that diagnoses its own mistakes and rewrites its operating rules raises hard questions about who actually sets policy.When automated systems repair relationship damage before a human even notices, accountability structures built around human review start to break down.This is The Judgment Layer, from Advisers Give Back — a nonprofit working to increase access to pro bono financial planning, pairing households who can't afford a financial planner with CFP professionals who volunteer their time. Each week, we take something that's actually happe Read the full transcript

  4. Sep 26

    The Overhang and the Butler

    The advisers who extract the most from AI are those with deep domain knowledge — not those with the most polished prompts.Your clients' AI agents may already be making financial decisions optimized for the agent's operator, not for your client.This is The Judgment Layer, from Advisers Give Back — a nonprofit working to increase access to pro bono financial planning, pairing households who can't afford a financial planner with CFP professionals who volunteer their time. Every episode starts with a real development in AI Read the full transcript

  5. Sep 22

    The Hollow Throughput

    When no one reads AI-generated work, an organization loses its grip on what it actually knows and has committed to.Speed and output volume can mask a quiet collapse in institutional judgment that stays invisible until something breaks.This is The Judgment Layer, from Advisers Give Back — a nonprofit working to increase access to pro bono financial planning, pairing households who can't afford a financial planner with CFP professionals who volunteer their time. Every episode starts with a real development in AI Read the full transcript

  • 5 Episodes

About

A twice-weekly show from Advisers Give Back on what AI is actually doing to the work of running a wealth management firm — client work, adviser development, operations, and what stays human. For the people who run mid-to-large RIAs.

Information

  • Creator
    Matt Iverson-Comelo
  • Years Active
    2026
  • Episodes
    5
  • Rating
    Clean
  • Show Website
    The Judgment Layer