A municipal greenhouse-plot tester can either attach twenty decorative site photos or paste the three posted rules as text when asking an LLM for cases. Which environmental choice is better when the text is sufficient?
Select an answer to reveal the explanation.
Short Explanation
If three posted rules already tell the story, hauling twenty decorative photos is like driving a truck to fetch a postcard. Lighter sufficient usage—text instead of image-heavy, few calls instead of many—is the better environmental choice. Model shopping for cost is a different conversation.
Full Explanation
Greenhouse-plot testers who can either attach twenty decorative site photos or paste three posted rules as text should prefer the lighter sufficient usage pattern—text over image-heavy, oversized, or repeated calls—when the text is enough. Driving a truck to fetch a postcard wastes energy.
Always attaching photos because multimodal input has zero energy cost fails by concept. Selecting a smaller SLM solely as a Chapter 5 cost strategy while ignoring usage pattern fails because this item targets usage pattern, not model shopping. Regenerating photo-heavy prompts repeatedly to “amortize” energy fails by increasing total consumption.
Exam caveat: lighter sufficient usage is the environmental choice when information content is already adequate in text.
Operational check: default greenhouse case generation to the three posted rules; allow photos only when visual defects are the oracle; cap attachment counts; reject decorative image spam in prompt reviews. Greenhouse prompt reviews should reject decorative photo attachments when the three posted rules already supply the decision basis, logging each multimodal exception with a rationale. Default templates to text-only so lighter sufficient usage is the path of least resistance.