Opening
AI for credit unions
Not a pitch
Images reveal affinity
Opening
The 3 questions
Everybody has AI
Buying labor
Code was deterministic
Not all AI is the same
Chat is not production
Why models get it wrong
Guardrails
Training is not one thing
Bias enters
Tokens are work
Right model, right job
Production grade
Context and format drift
Data does not disappear by magic
Training, fear, retention, risk
Controlled inference
Autonomy is a ladder
Agents need constraints
AI inside the perimeter
Blast radius
New supply chain risk
Board ROI
Where it fits today
Interrogate, do not extract
Stip clearing
Citations create trust
Fraud gets harder to see
Synthetic content examples
Tool conforms to user
Morning briefs
Protocol native ops
Questions to ask every vendor
Questions and feedback
← → arrow keys  ·  spacebar  ·  click to advance