🚩AI (as LLMs we know today) should not make decisions.
Also…
Machine Learning models such as regression, classification, forecasting etc. should not make decisions.
These techniques should ONLY provide insights to a decision model as inputs.
The explicitly defined decision model then uses those, as well as other inputs it needs to make a decision.
This is how you make a decision clear and transparent.
✨Decision models such as Decision Model and Notation (DMN) and Dynamic Decision Graph (DDG) allow you to present a holistic view of a business decision. These models share a common understanding of how decisions are made across a wide range of stakeholders.
⭕LLM ≠ Decision model
⭕Regression ≠ Decision model
⭕Classification ≠ Decision model
⭕Forecast ≠ Decision model
Particularly when clarity and transparency of decisions are important, you should not let anyone, human or non-human actors (Agentic AI) make any decision without following an explicitly modeled decision model.
How does that work?
1️⃣ Create a decision model (DMN or DDG)
2️⃣ At the conceptual level, specify the metrics of the nodes (i.e. decision units)
3️⃣ Associate the decision logic with those decision units (ML, LLM, Rules, etc.)
4️⃣ Define the relationship between the decision units.
The above 👆 steps will create a CompositeAI model that brings various decision logic components together to execute a decision model.
💣Decision model is not the decision logic you are familiar with, but that logic is definitely part of the holistic decision model you need to create.
–
Follow the “Uncle of DI” if you’re looking for unfiltered insights into DecisionIntelligence, AI, and DecisionAutomation. Hit the 🔔 on my profile to get notified about my daily posts.
Posted here.
Published October 3rd, 2025 at 07:30 am
