Understanding ChatGPT — What It Is and How It Works
Lesson 4: The Four-Tier Verification Framework — Applied
Lesson Objectives
By the end of this lesson, students should be able to:
- Apply the four-tier verification framework to a diverse range of real-world situations
- Use the consequential decision test to determine the appropriate tier
- Recognize the fluency trap in practice
- Identify the five non-negotiable professional review categories
Lesson Content
Quick review: the four tiers.
- Tier 1 â€" Low stakes: Plausibility check. Easy to correct if wrong. (Brainstorming, casual learning, organizing your thoughts)
- Tier 2 â€" Medium stakes: Spot-check important claims. (Client-facing outputs, information you will share publicly, decisions with meaningful cost)
- Tier 3 â€" High stakes: Verify all material claims. (Financial analysis, leadership presentations, professional documents)
- Tier 4 â€" Non-negotiable: Licensed professional review required regardless of output quality. (Medical, legal, financial investment, safety-critical)
The consequential decision test.
Before using any ChatGPT output: *"If this is wrong, what is the worst realistic outcome?"*
- Outcome: easily fixed, low cost â†' Tier 1
- Outcome: embarrassment, wasted time, meaningful cost â†' Tier 2
- Outcome: significant professional, financial, or reputational harm â†' Tier 3
- Outcome: health, legal liability, irreversible financial harm â†' Tier 4
Applying the test across common situations.
| Situation | Worst Case if Wrong | Tier | |———–|——————–|—-| | Brainstorming idea names for a project | Weak ideas â€" easy to discard | 1 | | Drafting an internal team email | Minor miscommunication | 1-2 | | Summary of a report for a client presentation | Client receives incorrect information | 2-3 | | Statistics cited in a document to leadership | Leadership makes a decision on bad data | 3 | | Legal contract language | Legal liability, contract disputes | 4 | | Medical dosage or treatment information | Patient harm | 4 |
The five non-negotiable professional review categories.
These five categories are always Tier 4 â€" no exceptions, regardless of ChatGPT's output quality:
- Legal decisions: Contract terms, regulatory compliance, legal strategy â†' licensed attorney
- Medical decisions: Diagnosis, treatment, medication â†' licensed physician or appropriate provider
- Financial investment advice: Investment strategy, tax planning, retirement â†' licensed financial advisor or CPA
- HR and employment decisions: Termination, accommodation, classification â†' HR expertise + legal counsel
- Safety-critical engineering: Structural designs, safety systems â†' licensed engineer
The fluency trap in practice.
Recognizing it: ChatGPT produces a response with a precise-sounding statistic, a confident recommendation, or a detailed explanation. It is well-written, well-organized, and reads authoritatively. Without independent verification, the quality of the writing creates an impression of accuracy.
Defending against it: Ask "how would I verify this?" before acting on any specific factual claim. If you cannot describe how to verify it, that is your signal to find out how before acting.
Enroll to continue this lesson.
The preview above shows the lesson objectives and opening lesson content. Enroll to view the full lesson, complete the practice work, and take the lesson quiz.