# Designing With & For AI

**Track:** Systems, Strategy & Frontier — Design & Human-Centered Design — proposed (25)
**Framework / surface:** design
**Level:** Advanced
**Prerequisites:** Human-Centered Design Foundations, Nielsen’s Usability Heuristics

**In one line:** Human-AI interaction patterns, generative UX, and calibrating trust under uncertainty.

## Theory, aesthetics & inspiration

Designing for AI means designing for probability: a generative system answers the same prompt differently each time, so the interface must make uncertainty legible rather than conceal it. The discipline now rests on codified guidance — Microsoft's eighteen Guidelines for Human-AI Interaction (Amershi et al., CHI 2019), operationalized in the HAX Toolkit (2021), and Google PAIR's People + AI Guidebook (2019). Beneath both lies Lee and See's "Trust in Automation" (2004), which named the real target: not maximal trust but calibrated trust, reliance matched to actual capability. Hence current generative-UI practice — and NIST's AI RMF Generative AI Profile (2024) — treats error, control, and explanation as first-class material, not garnish.

## References

- [People + AI Guidebook — Google PAIR](https://pair.withgoogle.com/guidebook/)
- [Microsoft HAX Toolkit (Guidelines for Human-AI Interaction)](https://www.microsoft.com/en-us/haxtoolkit/)
- [Artificial Intelligence Risk Management Framework: Generative AI Profile (NIST AI 600-1)](https://www.nist.gov/publications/artificial-intelligence-risk-management-framework-generative-artificial-intelligence)