Constrained Decoding for AI-Generated Podcast Tags
Daniel's AI-generated show deploys to Vercel, runs on serverless Postgres, and produces exactly the kind of wide-ranging content he wants. The problem is the feed — automatic tagging that started inventing its own categories, then duplicating them in ways nobody notices. This episode works through the fix: why unconstrained categorization never converges, how JSON Schema enums and context-free grammars make label bloat structurally impossible, and why a fine-tuned encoder beats a frontier model for a stable label set at one to two orders of magnitude lower cost. We also cover the…
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.