daBongo LMS AI Training Courses

Lesson Content

Why "please review this" does not work.

Vague review requests produce vague feedback: "This is well-written and professional. Consider strengthening the conclusion." This is useless.

Useful review requires criteria – specific standards against which the work is evaluated:

"Review this [document type] against these specific criteria: (1) [criterion 1], (2) [criterion 2], (3) [criterion 3]. For each criterion, tell me: does this document meet the standard? Where specifically does it fall short? What would fix it?"

The audience-perspective review.

The most important question for any communication is: does this land with the intended reader?

"Read this [document] as a [specific reader type]. What would be your first impression? What question does this leave unanswered that you would want answered? Where would you stop reading and why? What would make you take the action this piece is asking for?"

The "weakest element" test.

"If you had to identify the single weakest element of this piece – the thing that most undermines its effectiveness – what would it be? Why? How would you fix it?"

The "single weakest element" constraint forces prioritization rather than a list of everything that could be improved.

The "what would a skeptic say" review.

For persuasive content:

"Read this [proposal / argument / recommendation] as a well-informed skeptic who is looking for reasons to push back. What are the three strongest objections a skeptic would raise? Where is my argument most vulnerable? What additional evidence or reasoning would address each objection?"

Limit of AI review: factual and professional accuracy.

Meta AI catches logic, structure, tone, and clarity problems well. It cannot reliably catch:

• Factual errors about real-world specifics
• Legal, regulatory, or compliance accuracy
• Technical accuracy in specialized domains
• Whether professional claims are accurate and appropriate

For any deliverable where factual, legal, technical, or professional accuracy matters, human expert review is always required regardless of Meta AI's assessment.