brainstorming-research-ideas

Use structured ideation frameworks rather than waiting for inspiration. Start with tension hunting to find trade-offs in your field, check what has recently changed in compute or data, then filter candidates with the two-sentence explain-it test and feasibility checks.

Begin with the tension and contradiction hunting framework to list trade-offs in the field, then apply the what-changed principle to find recently invalidated assumptions. Generate 10-20 candidates before filtering down to the strongest 3-5 ideas.

Skip ideation frameworks when you already have a well-defined research question and need execution or experimental design guidance. They are also not a substitute for a literature review, which requires a dedicated literature-review skill.

Apply the converge-phase filters: state the idea in two sentences, confirm a genuine problem exists, verify a simpler approach does not suffice, and identify a clear beneficiary. Then define a two-week pilot experiment to test feasibility.

Common pitfalls include novelty without impact, incremental benchmark-chasing, unnecessary complexity, echo chambers from reading the same papers, and premature convergence on the first idea. Each pitfall maps to a specific corrective framework.