Designing for equivalent experiences is a critical goal for developers building accessible AI agents. Aiming for an equivalent experience across all abilities helps all users achieve the same goals and derive comparable value, even if the specific interaction methods differ based on the context or need.
Designing for equivalent experiences involves:
- Considering the different ways people might interact with the software, such as using touchscreens, captions, keyboards, or voice commands.
- Helping interfaces function effectively with different input and output modalities.
- Helping all users find and use interactive elements with relative ease.
A useful model for developers involves designing the agent interface to deliver equivalent input and output experiences.
- Input equivalence: enable any user to provide inputs effectively, regardless of their preferred or required modality.
- Output equivalence: enable any user to receive outputs effectively, helping to enable comparable information transfer regardless of their preferred or required modality.
The ultimate goal is for all users, regardless of their abilities, to have the same level of satisfaction when using the product, and to have maximum control over how they interact with it.