Preserve the pre-formal layer of human intellectual work.
Real conversations generate what we call prayers — implicit intentions, half-formed intuitions, gestured-at connections that are too vague to write down but too important to lose.
A theory builder saying "I keep noticing this pattern across these three examples, I don't know what it is yet" is not wasting time. That gesture is the actual creative work. The formalization comes later. The insight comes here.
Current tools handle this badly:
- The insight evaporates when the conversation moves on
- Note-taking interrupts the social rhythm
- Linear transcripts lose the structure of what was developing
- No system tracks how a vague intuition accumulates specificity across sessions
Threads is infrastructure for the live selection layer — the part of the knowledge stack where human creative direction originates.
Autoformalization and verified software engineering are collapsing the cost of producing formal knowledge. The scarce resource is shifting to specification — human taste, judgment, and direction. Specifications are set in conversations.
Threads makes the conversational layer legible to the formal infrastructure without forcing it to be formal.
Layer 0 CONVERSATION Pre-formal, gestural, exploratory.
Prayers emerge here.
↕
Layer 1 THREADS Captures prayers with context.
Tracks threads across sessions.
Surfaces connections, cruxes, lulls.
↕
Layer 2 JUST-IN-TIME FORMALISM Offers candidate formal statements
for human review when ready.
↕
Layer 3 FORMAL BACKBONE Verification is now cheap.
This layer is being built by others.
↕
Layer 4 FEEDBACK Verified signals flow back to the conversation.
Threads listens to your conversation and builds a live graph — threads, claims, cruxes, tangents — without interrupting the flow.
Core capabilities:
- Prayer detection — identifies pre-formal intentions before they evaporate, preserves them with full conversational context
- Thread tracking — open threads persist across sessions with surrounding context intact
- Claim decomposition — factual, normative, and worldview claims surfaced and distinguished
- Crux visibility — what agreement depends on, where conflict roots are
- Retrieval at lulls — surfaces dormant threads at natural pauses with suggested re-entry phrasing
- Speaker analytics — speaking time, turn-taking balance, bandwidth distribution
- Rhetorical pattern detection — fallacies and rhetorical moves flagged with confidence scores
- Formalization bridge — when a prayer has accumulated enough context, offers a candidate formal statement for human review
Conversation is primary, tools are substrate. No feature should force participants to leave conversational mode. Threads is infrastructure, not an interlocutor.
Preserve specificity, resist abstraction. A prayer captured with full context is more valuable than a generalized summary. Specificity is the raw material of eventual formalization.
Offer, never direct. Every intervention is an offering — surfaced threads, suggested connections, formalization prompts — all optional, all human-confirmed.
Transcript as source of truth. Every analysis is traceable back to concrete utterances.
Privacy-first. Local-first inference where feasible. Self-hostable. Your data stays yours.
The formal backbone of knowledge is being built right now. Specification decisions made today will become load-bearing infrastructure — socially and epistemically expensive to change, even when computationally cheap to regenerate.
The governance layer that keeps human judgment in the loop at the point where creative direction originates needs to develop in parallel. The mould is being set.
Threads is part of that governance layer.
Backend: FastAPI (Python), PostgreSQL, local-first LLM routing (LM Studio / Ollama), Whisper STT
Frontend: React + Vite, React Flow
AI: Local-first by default (privacy-preserving), optional cloud LLM fallback
Conversations generate structure. Structure generates knowledge. Knowledge changes what is possible.