A historic-mill gift shop likes a demo voice today and already knows it will later teach a house style from its own catalog notes. Which selection filter should the team apply now?
Select an answer to reveal the explanation.
Short Explanation
Think of a gift shop that likes today's demo and already knows it will teach a house style later. Customization has to be on the first pick. A pretty dashboard does not create that path.
Full Explanation
The historic-mill gift shop already likes today’s demo voice and already knows it will later teach a house style from its own catalog notes. That future customization need must shape the first selection: customization availability is the filter to apply now. Choosing a model that cannot be customized locks the shop into a dead-end even if the first demo sounds charming, because house-style teaching later requires a supported customization path on the same foundation choice.
Selecting on first-demo charm only fails because a pleasant sample does not create a later path to teach catalog-note style. Picking a model with no customization options deliberately fails the stated roadmap: if the voice can never change through customization, the planned house style cannot be applied. Ignoring customization and picking by which dashboard draws the prettiest chart fails because visualization tooling is unrelated to whether a generative model supports customization.
Exam caveat: model selection is not only about today’s demo quality; practitioner designs that foresee fine-tuning, continued pretraining, or other customization must verify that path exists before committing. Operational check: before approving the gift-shop model, confirm the customization mechanisms available for that model family, sketch how catalog notes would be applied later, and reject any shortlist item that only wins on demo charm.