Question 1

But GPT-5 already understands vague prompts — why train people?

Accepted Answer

Because the failure mode has shifted. Older models failed visibly ("I don't understand"). Frontier models fail invisibly — they answer confidently and plausibly even when wrong. The bottleneck is no longer phrasing; it's the human's Clarity to scope the problem, the Context they supply, and the Confidence to know what to trust. That's a skill, and it compounds.

Question 2

Will Keypra become obsolete when models get smarter?

Accepted Answer

The opposite. Smarter models reward sharper thinkers. As capability grows, the gap between professionals who can deploy AI well and those who can't widens — not narrows. Keypra trains the Human Layer, which is the part the model can't replace.

Question 3

Isn't this just compliance theatre?

Accepted Answer

EU AI Act Article 4 (in force February 2025) requires "a sufficient level of AI literacy" for staff using AI systems. That is a legal duty, not a recommendation, and regulators expect evidence. Keypra produces that evidence — per-user training records, competence scores, and date-stamped audit trails — so compliance teams have something defensible to hand to a DPA.

Question 4

What do you mean by "the Human Layer"?

Accepted Answer

It's everything an AI model still can't do for you: framing the right question (Clarity), giving it the knowledge it doesn't have (Context), and deciding what to trust before it leaves your desk (Confidence). Keypra builds those three habits through practice, not theory.

Question 5

What does the EU AI Act actually require?

Accepted Answer

Article 4 requires providers and deployers of AI to ensure their staff have "a sufficient level of AI literacy", taking into account technical knowledge, experience, education, and the context the AI is used in. Article 13 requires transparency for high-risk systems. Keypra's AI Literacy Methodology is mapped directly to both — read the full note in our compliance section.

Question 6

Why use Keypra when AI models keep getting smarter?

Accepted Answer

Smarter models don't replace your thinking — they amplify it. Frontier models like GPT-5 and Claude Opus 4.5 still produce confident-but-wrong answers, still need precise context to be useful at work, and still leave the human accountable for the output. Keypra trains the Human Layer of AI at work — Clarity, Context, and Confidence — and under the EU AI Act Article 4 ensuring this AI literacy across your workforce is now a legal duty for European employers.

Shortcut	Action
`Cmd/Ctrl + K`	Open command palette / search docs
`?`	Show in-app shortcut reference
`G then H`	Go to Home
`G then L`	Go to Practice Lab
`G then C`	Go to Composer
`G then P`	Go to Palace

Shortcut	Action
`Cmd/Ctrl + S`	Force save (auto-save also runs continuously)
`Cmd/Ctrl + Z / Shift + Z`	Undo / Redo
`Cmd/Ctrl + B / I`	Bold / Italic
`@`	Insert a persona
`[[`	Insert a wikilink to a document
`![[`	Embed a block from another document

Shortcut	Action
`Cmd/Ctrl + Enter`	Copy composed prompt
`Cmd/Ctrl + Shift + L`	Send current prompt to Practice Lab
`Cmd/Ctrl + Shift + S`	Run Prompt Stress Test
`Cmd/Ctrl + Shift + E`	Export composition

Shortcut	Action
`Cmd/Ctrl + Enter`	Run analysis on current prompt
`Cmd/Ctrl + Shift + R`	Re-run with the same model

Shortcut	Action
`→ / ←`	Next / previous slide (lecturer)
`R`	Re-sync with the lecturer (learner)
`F`	Toggle full-screen

Keyboard shortcuts

Global

Editor (Documents, Cards, Personas)

Composer

Practice Lab

Classroom (Rafy)

Palace map / Research Canvas

Shortcut	Action
`F`	Toggle Focus mode
`/`	Focus the search box
`Space + drag`	Pan the canvas
`Cmd/Ctrl + scroll`	Zoom