A hardware retailer wants to add a search feature that lets customers type a plain-language question like 'a drill that works well for tile and concrete' and get relevant product matches, even when the exact words do not appear in the product descriptions. The retailer is evaluating whether this requires generative AI or a simpler approach. What should guide that decision?
Select an answer to reveal the explanation.
Short Explanation
You don't need the heaviest, most expensive tool here. Semantic search can match meaning and intent beyond exact keywords, so a query like 'a drill that works well for tile and concrete' can surface a hammer drill even if those words aren't in the listing. Plain keyword matching misses that, and there's no multi-turn conversation to justify generative AI.
Full Explanation
The correct choice is the semantic search option, which does not require generative AI. Semantic search is built to understand the meaning and intent behind a query, matching a question like a drill for tile and concrete to relevant products even when those exact words are absent from the descriptions, and it does this through vector-based similarity rather than needing a generative model to write new text. The option claiming generative AI is required because only large language models can process natural language questions is incorrect because processing natural language questions does not require a generative model specifically, since semantic search and other language understanding techniques can interpret intent without generating novel text. The option saying a simpler keyword search option is sufficient because it already matches exact words in descriptions is incorrect because plain keyword search only matches literal words present in the text, which is the exact limitation the retailer is trying to overcome. The option asserting generative AI must be used because product search always involves multi-turn conversation is incorrect because this scenario describes a single-turn product lookup, not an extended back-and-forth conversation, so there is no requirement for multi-turn dialogue handling here.