Courses · Method
Eight principles from learning science — each one is something you do in a lesson.
They are not decoration. Every principle below corresponds to a part of the course you will feel while taking it, and each names the research it comes from.
- 01
Metacognition core
Externalize thinking BEFORE you type a prompt.
Every step forces you to write your reasoning first — what you actually need, who you're talking to, what "good" looks like. The real prompting skill is metacognition, not the words you ultimately use. Research on expert problem-solving shows that the time to think about the problem dwarfs the time to solve it.
Newell & Simon, Human Problem Solving, 1972
- 02
Deliberate practice + choice-based learning
Pick between vague vs structured paths — see the consequences in real time.
Every lesson presents A/B thinking-path choices: "just hit enter" vs "front-load the structure." The system reveals immediate, specific consequences — wasted tokens, mediocre outputs, lost time. Deliberate practice with feedback is how skills are built; watching is not (Ericsson et al., 1993).
Ericsson, The Role of Deliberate Practice, 1993
- 03
Dual-coding & visual transformation
See the bad prompt → see the good prompt → spot every structural key.
Side-by-side animated bad vs good prompts. Glowing "WHERE IS THE KEY" hover callouts mark the six structural ingredients every elite prompt has: role, chain-of-thought, delimiters, output schema, few-shot examples, constraints. Seeing the structure and reading it together is remembered better than either alone (Mayer's multimedia principle).
Paivio, Mental Representations, 1986
- 04
Scaffolding & cognitive apprenticeship
Start simple, ramp complexity, hear the expert think out loud.
Lessons begin with one variable changed at a time. She, your AI guide, narrates the expert thinking process in real time — the same way a master shows an apprentice. Complexity is added gradually, in chunks small enough for working memory.
Collins, Brown & Newman, Cognitive Apprenticeship, 1989
- 05
Active recall + instant feedback
Type your answer. Get a specific critique. Retry.
After every example or challenge, you get instant grading on both your thinking AND your prompt — never "nice job" without specifics. Growth-mindset language: "missing audience targeting" not "wrong." Retry is encouraged, not shamed. Being tested on something makes it stick far better than rereading it (Roediger & Karpicke, 2006).
Roediger & Karpicke, The Testing Effect, 2006
- 06
Pre-AI thinking drills
Show your full thinking process before you write a single word of prompt.
Every mini-test starts with: "Before you ask any AI anything, show your full thinking process." The grader critiques the reasoning quality separately from the prompt quality. This builds the exact mental model you'll use every day with the AI you already use — and stops you from spamming the model with half-formed asks.
- 07
Flow state & dopamine design
Short bursts. Confetti on correct choices. Streaks. Brain Upgraded.
60–90 second chunks. Confetti on correct picks. Progress streak counters. Subtle teal glow + sparkle on "aha!" moments. A celebratory "Brain Upgraded" screen with shareable certificate at the end. Short chunks with a clear next step keep you going to the end.
Csíkszentmihályi, Flow, 1990
- 08
Adaptive personalization via She
Your AI guide picks the right path — and keeps nudging you forward.
She asks 5–7 questions about goals, pain points, and experience, then maps a personalized free → paid path. She also points you to the next course that fits: "You lean on instructions a lot — Output Formatting Secrets unlocks the next step in 6 minutes."
36 courses · 36 free
Start with a learning path — the order is done for you, and a milestone counts when you pass its quiz.