Blog

26th MAY 2026

Human-to-AI Interfaces (HAII)#

HAII is to AI what HCI is to computers. A sister discipline with its own constraints, affordances, and failure modes, and a manifesto for treating it that way.

Read more →
22nd MAY 2026

Inline Agents#

An LLM that edits the user's draft continuously, in place, on every pause-in-typing. The continuous extension of Inline Prompting.

Read more →
18th MAY 2026

Inline Prompting (Blanks)#

Inline Prompting allows the user to request LLM assistance by placing a _ in any text input. The user→LLM cue of the OpenCues two-direction model.

Read more →
14th MAY 2026

Inline Cues (Continuous Onboarding)#

Inline Cues are system-to-user signals that teach users about the tool in context, while they use it. The structural answer to AI-era feature churn outpacing documentation.

Read more →
10th MAY 2026

Human Interaction#

Human conversation is faux turn-based, with constant non-blocking cues. Current LLM chat is strictly turn-based, with none.

Read more →
6th MAY 2026

HCI (Human to Computer Interface)#

HCI is any means through which we interface with a computer. The sharpest framework I know for evaluating one is borrowed from fighting games.

Read more →