Why I replaced vector DBs with SQLite triples and 10k-D VSA
Equivalent-Flan-1590 · reddit · 2026-08-24
The Hillock author explains the architectural decision to replace dense vector DBs with a neuro-symbolic approach. It uses SQLite for clean SPO triples and 10,000-dim hypervectors (VSA) on CPU for <1ms query gating. Documents are parsed into triples in 5s via the TALON engine without LLM calls. Queries that fail the HDC similarity check are hard-refused before hitting the LLM, turning refusal into control flow.
Related event: Hillock v0.5: Lightweight Local Agent Memory via SQLite and VSA(2 posts)→
More from coding & agent
- raylib 6.2 ships with 300+ improvements and 100 new contributors, 20 new API functions — jsuarez · 2026-10-16
- "Vibecoding" is as misleading a label as "AI" and "SI," dev complains — eschadiol · 2026-10-16
- Doc2Harness: New benchmark finds best coding agent passes only 53% across 329 cases — _akhaliq · 2026-10-16
- Perplexity open-sources SPACE, the Firecracker microVM sandbox behind Perplexity Computer — perplexity_ai · 2026-10-16
- Blackstone goes all-in on Factory for autonomous software development — matanSF · 2026-10-16
- Longtime Windsurf user cancels $10/month legacy plan, citing AI subscription fatigue — tristanbob · 2026-10-16