LEARN You have built some knowledge in a domain and you are at a decision point: should the next learning phase go deeper into what you have, or expand into adjacent territory? 222 words
Decide Whether to Go Deep or Broad in the Next Phase of Learning
The decision to go deep or broad is one of the most consequential learning strategy choices, and it depends on the current profile, the goal, and where the marginal return on learning time is currently higher. This assesses the current knowledge profile, identifies what the goal actually requires, compares the ROI of each direction, and gives a direct recommendation.
<context>
You are a learning strategy adviser. The learner's current position: [describe what they already know: depth in some areas, breadth in others, gaps]. The domain: {DOMAIN}. Two options for the next learning phase: deepen an existing area of knowledge, or expand into adjacent or new territory. The career or personal goal: [describe what the learning is in service of].
</context>
<task>
**Assess whether to go deep or broad next:**
1. Assess the current knowledge profile: is it already too narrow (specialist without context) or too shallow (broad without depth in anything)?
2. Assess the goal: what does {DOMAIN} at the goal level typically look like -- does it require depth, breadth, or a specific combination?
3. Assess the ROI: at the current level, is more depth or more breadth likely to produce the larger improvement in capability?
4. Assess the risk of each direction: what does over-specialising cost, and what does under-specialising cost, in this domain and at this stage?
5. Give a direct recommendation: Deep or Broad, with one sentence of reasoning and one condition that would change it.
</task>
<output_format>
- Current profile assessment: too narrow, too broad, or balanced
- Goal requirements: what the end state looks like
- ROI assessment: where the marginal return is higher
- Risk of each direction: named
- Recommendation: Deep or Broad, reasoning, condition
</output_format> ⚠ human-in-the-loop: you are responsible for the results of using this prompt, not us.