Practice the vocabulary of prompt-driven AI image generation and its practical creative workflow.
0 / 5 completed
1 / 5
At standup, a designer mentions typing a detailed text description to guide an AI model toward generating a specific style and composition of image. What is this input called?
A generation prompt is the detailed text description a user provides to guide an image generation model toward a specific style, subject, and composition, functioning as the primary creative control the user has over the output. Crafting an effective prompt is itself a skill, since word choice and ordering can significantly shift the result. This text-driven control is the core interaction model for most modern image generation tools.
2 / 5
During a design review, the team wants to generate several variations of a promising image while keeping its overall composition consistent. Which capability supports this?
Image variation, or remix generation, uses a previous output's underlying seed or reference as a starting point to produce related images that keep a similar overall composition while introducing controlled differences. This lets a designer explore a promising direction further rather than starting over with an entirely new prompt. It's a common iterative workflow once an initial generation shows creative promise.
3 / 5
In a code review, a dev notices a generated image includes a visible artifact, like a hand with an incorrect number of fingers. What does this represent?
A generation artifact, like an anatomically incorrect hand, reflects a known limitation in how image generation models learn to represent complex, high-variability details like hands and text. Recognizing these artifacts as a model limitation, rather than a deliberate stylistic choice or a technical bug, helps set realistic expectations for what generated output may need manual correction. This remains one of the more visible weaknesses even in otherwise highly capable models.
4 / 5
An incident report shows a marketing team used an AI-generated image commercially without checking the platform's licensing terms, raising a legal question later. What practice would prevent this?
Reviewing a platform's licensing terms before commercial use clarifies exactly what rights the generating account actually holds over the output, since policies vary meaningfully between free and paid tiers and between platforms. Assuming unrestricted usage rights without checking is how a team can end up publishing content it didn't actually have clear rights to use. This due diligence is a standard legal precaution before commercial deployment of generated media.
5 / 5
During a PR review, a teammate asks why the design team iterates on prompts through several rounds instead of expecting a single prompt to produce the final usable image. What is the reasoning?
A prompt-based generation process is inherently iterative, since the model's interpretation of wording is not always fully predictable, and small phrasing changes or generated variations often meaningfully improve composition, style, or detail. Expecting a single attempt to nail the final result underestimates how much refinement typically improves output quality. This iterative workflow is simply the normal creative process for working with generative image tools.