Philosophical
Research System
A multi-agent AI platform built for objectivity: it maps the strongest opposing positions on any philosophical question, argues each at full strength, and forces the answer to survive blind critique, scored evaluation, and its own devil's advocate before it is written.
How It Works
The user types any philosophical question and picks intellectual frameworks from a catalog of 15 methodologies. A six-stage pipeline then runs — advocates in parallel, everything else in sequence — streaming its progress live, so you watch the positions form, clash, get scored, and converge.
Position Mapper
Reads the question and maps the strongest genuinely opposing positions it allows — the axis the whole pipeline will argue across.
Steelman Advocates
One advocate per position, running in parallel. Each argues its side at full strength under strict citation rules — and must name the weakest point of its own case.
Blind Critic
Critiques every brief without being told which framework or position produced it, hunting for framing bias, missing voices, and shared blind spots across all sides.
Scored Rubric
A structured evaluation forced into valid JSON: argument quality, evidence discipline, and citation verification — suspect references get flagged, not repeated.
Devil's-Advocate Refuter
Attacks the draft synthesis before it settles, so the conclusion has already survived its strongest objection.
Final Answer
The revised conclusion — ending with an explicit 'epistemic status' section that states how confident the answer deserves to be, and why.
Framework Catalog
Technical Stack
What I Learned
Most importantly, this project confirmed my core belief: AI is most powerful not when it replaces thinking, but when it structures it, giving different frameworks equal weight, holding contradictions open, and refusing easy synthesis.