Anyone who works with AI tools knows the frustration of getting responses that don’t quite hit the mark. You craft a prompt, get an answer, and then realize it’s not quite what you needed. This back-and-forth can be exhausting and make it hard to trust the AI. To help fix this, a new framework called Promptions has been created to give users more control over AI outputs through dynamic UI options.

What Are Promptions and How Do They Work?

Promptions is a simple yet powerful UI framework that helps developers build AI interfaces with more precise user controls. Instead of just typing a prompt and hoping for the best, users can choose from options that refine the AI’s response. These options can include things like response length, tone, or starting point. The goal is to make interactions more predictable and tailored to each user’s needs.

This framework is easy to integrate into existing AI systems. It’s especially useful in areas where context and nuance matter, such as customer support, education, or healthcare. By giving users a way to adjust parameters on the fly, Promptions makes AI interactions smoother and more effective, reducing the frustration of long, trial-and-error exchanges.

The Research Behind Promptions

The idea for Promptions comes from research into how people use AI when their goal is understanding rather than creating. The researchers focused on tasks like explaining complex formulas or code snippets, which can vary greatly depending on who’s asking. For example, a beginner might need a simple breakdown, while an expert might want a detailed debugging guide.

The research found that users often struggle to phrase questions in a way that gives the AI the right amount of detail. They spend a lot of time tweaking prompts or managing the interaction instead of actually learning or understanding the material. This led to the idea of adding controls that let users guide the AI more precisely, based on their specific needs and context.

Designing Better AI Interactions

To explore how users want to steer AI outputs, researchers conducted studies with professionals from various fields, including engineering, marketing, and research. Participants reviewed mock-ups of static prompt options—like adjusting tone or response length—but found these were helpful yet limited. Static options couldn’t adapt to different types of content, such as code snippets or detailed explanations.

This feedback helped shape Promptions into a more dynamic system. Instead of fixed choices, the framework allows for real-time adjustments that are more context-aware. This way, users don’t have to guess what prompt settings work best—they can directly control the AI’s behavior in a more natural and flexible way.

Overall, Promptions aims to make AI interactions more intuitive and tailored. By embedding simple controls into AI interfaces, it helps users get more accurate and relevant responses, saving time and reducing frustration. It’s a step toward making AI tools more user-friendly and adaptable across different fields and tasks.

Inspired by

• https://www.microsoft.com/en-us/research/blog/promptions-helps-make-ai-prompting-more-precise-with-dynamic-ui-controls/

Sources

• azure.com