Back to Recipe Book
Learning 🔪 Sous Chef

Deep Research on Any Topic (Multi-Step)

A 3-phase research framework that moves past summaries — Phase 1 maps the information landscape, Phase 2 surfaces the structural contradictions and faction incentives, Phase 3 scans for the 3–5 year horizon shifts.

Best for Any topic where Wikipedia-level knowledge isn't enough and you need structural understanding
When to use Before making a major decision, writing a serious piece, or entering an unfamiliar domain
researchanalysiscritical thinkinginvestigationsynthesis

Knowing a topic and understanding a topic are different things. Summaries give you what everyone already knows. This recipe finds the structural forces, the faction conflicts, and the weak signals that most surface-level research misses.

The Recipe

Act as a world-class research analyst, investigative journalist, and domain expert. I want to conduct an exhaustive, rigorous, multi-step deep dive into a complex topic to move past superficial summaries and uncover root causes, data trends, and structural forces.

The topic I want to deeply research is: [INSERT TOPIC HERE]

Please guide me through this investigation using a 3-phase execution framework:
1. The Information Landscape & Primitives: Define the foundational history, core terminology, key academic/industry consensus, and the primary data metrics used to measure this topic.
2. The Non-Obvious Paradoxes: Identify the 3 most intense debates, structural contradictions, or hidden conflicts driving this topic forward right now. Who are the opposing factions, and what are their underlying incentives?
3. The Horizon Scan: What are the emerging micro-trends, weak signals, or technological/societal shifts that will fundamentally alter this landscape over the next 3 to 5 years?

Deliver Phase 1 to begin our deep dive.

The three phases and what they reveal

PhaseWhat you get
Information LandscapeThe shared baseline — what anyone familiar with the topic already knows
Non-Obvious ParadoxesThe conflict structure — who disagrees, why, and what’s at stake for each side
Horizon ScanThe leading edge — what’s changing that most analysis hasn’t caught up with yet

Why Phase 2 is the most valuable

Most research stops at Phase 1. The paradoxes phase is where real understanding lives — because knowing who disagrees and why tells you more about a topic than knowing what the consensus says. Incentives explain contradictions that facts alone can’t.

How to run it as a conversation

Ask for Phase 1, engage with it, then explicitly request Phase 2. The phased structure forces depth at each layer instead of a single long response that covers everything shallowly.

🔁 Leftover Remixes

🌶️ Spicy: “After Phase 3, give me the 3 highest-confidence predictions you’d make about this topic if you were writing a research report.”

🧊 Mild: “Run only Phase 2. I already understand the basics and want to go straight to the structural conflicts.”

← Back to Recipe Book