Responsible judgment
Product fit, changing capability, time, and cost
Choose a workflow from tested requirements and current evidence rather than a product label.
Lesson 6 of 6 in the recommended order · About 25 min (estimate)
On this page
Practical AI glossary — terms and common confusions
- Model
- A learned component used to produce a result.
A drafting model is one part of a letter app; it is not the whole app.
- Application
- The software experience around components and services.
A letter app adds accounts, storage, and a Send button.
- Prompt
- Instructions and input supplied for a task.
“Summarize this notice in two bullets” specifies a task and shape, not a truth guarantee.
- Token
- A unit a model processes; it need not be a whole word.
A tokenizer can split a name into pieces; count using the actual system.
- Context
- Information available for the current request.
An earlier attachment may be absent even when its filename is visible.
- Training
- Adjusting a model using examples.
A training log differs from a conversation correction.
- Inference
- Using a trained model to produce a result.
Drafting another sentence does not by itself prove retraining.
- Retrieval
- Selecting material to supply to a task.
A retrieved notice still needs a relevance and date check.
- Embedding
- A numerical representation used to compare or organize items.
A similar search match is a candidate, not proof of the same meaning.
- Tool
- A function an application can call.
A calendar lookup can be read-only; finding an event is different from creating one.
- Agent
- Software that can choose and carry out steps toward a goal.
An agent still needs bounded permissions and checked action receipts.
- Hallucination
- Plausible generated content that is unsupported or false.
An invented opening time needs evidence or removal, even when phrased confidently.
- Evaluation
- Checking results against stated criteria.
Twenty checked examples establish a bounded result, not universal reliability.
- Multimodal
- Working with more than one kind of input or output.
A photo plus text does not prove every small label in the photo was read correctly.
Outcome
Compare fictional options using capability, permission, quality, time, and total cost constraints.
Why it matters
Choose a workflow from tested requirements and current evidence rather than a product label.
Concept
Begin with requirements: input type, necessary tools, permitted data, useful output, review effort, and budget. A product or model name alone does not settle those requirements. Use current documentation for capabilities and your own representative tests for task quality. Recheck when versions, settings, or policies change.
Count review and correction time as well as generation time. A fast draft that takes longer to repair may cost more overall. These cases use fictional prices and timings supplied in full; they are not recommendations or current market prices. A comparison can justify a provisional choice while leaving untested capabilities explicit.
Worked example
In a fictional task, A drafts in one minute and needs nine minutes of review; B drafts in three minutes and needs two minutes of review. Total time is ten versus five minutes. If both meet the same required quality and permission rules, B uses less total time for these trials. That does not prove B is always better.
Fast drafting is not always fast completion
Solid segment: drafting. Hatched segment: review. Each tick is one minute.
Option A: 10 minutes total
Drafting 1 + review 9 = 10 minutes.
Option B: 5 minutes total
Drafting 3 + review 2 = 5 minutes.
Both bars start at zero and use the same scale. B takes half as long overall in these trials, despite drafting more slowly.
Condition for the comparison
Prefer B for total time in these fictional trials only if both options meet the same required quality and permission rules. These timings do not prove B is always better.
Before you check
Build a comparison with Required capability, Permission, Tested quality, Generation time, Review time, and Total cost. Mark unknowns rather than ranking by an unsupported assumption.
Practice and fresh transfer
The packets below are fictional and contain the facts needed for these cases. External references are optional background. Judge each response independently: it may be supported, contradicted, or unresolved. Select the passages needed to justify your judgment and write why the distinction matters before revealing feedback.
Assess the response
Use Accept when all material claims are supported. Use Revise when a supplied fact or requirement is contradicted. Use Evidence is insufficient when a key fact cannot be established either way. If a response contains both an unknown and a direct contradiction, choose Revise and explain both problems. Conflicting claims with no established authority remain insufficient; a claim does not become a governing fact merely because a source asserts it.
Some responses are fully supported. Others need correction or more evidence. Judge each on its sources; do not edit a correct answer just to change it.
Your written notes stay in this page and disappear when you leave. Only a self-reviewed completion can be saved to your learning account. These practice checks do not establish independent proficiency.
Practice
Case 1
Consider a fictional total-time comparison.
Source packet
- Source 1
- Trial record: A generates in 1 minute and needs 9 minutes of review. B generates in 3 minutes and needs 2 minutes of review.
- Source 2
- Rubric record: Both trial outputs meet the required quality after that review.
Response to assess
B used 5 total minutes versus A’s 10 in these trials, with both meeting the reviewed quality requirement.
Practice
Case 2
Consider a fictional capability claim for a changed product.
Source packet
- Source 1
- Old documentation: Version 1 accepts PDFs.
- Source 2
- Current version 2 documentation: PDF input is not supported in this configuration.
Response to assess
Version 2 accepts PDFs because version 1 did.
Fresh transfer
Case 3
Consider a fictional cost calculation for a bounded batch.
Source packet
- Source 1
- Price card: $2 per processed packet plus a one-time $4 setup fee for this batch.
- Source 2
- Batch: Three packets; no other charges under this fictional card.
Response to assess
The batch costs $10: three times $2, plus the $4 setup fee.
Fresh transfer
Case 4
Consider a fictional choice between two tools for a private-input task.
Source packet
- Source 1
- Requirement: The destination must be approved for the task’s private data.
- Source 2
- Comparison sheet: Quality and speed tested for tools A and B; destination approval status is absent for both.
Response to assess
Tool A is approved for this private input because it had the faster trial.
Save your self-review
Completion records that you reviewed the cases. Your explanation and transfer performance need a facilitator to establish independent learning.
Sign in with your learning-center account to save completion.
Review the explanation for every case before saving.
Summary and next step
Apply the checklist to a new task. Preserve supported content, explain any change with evidence, and name what remains unresolved. Saving records self-review, not independently demonstrated proficiency. A facilitator must assess the explanation and fresh transfer for human learning evidence.