4A — Rejected Name Candidates
Full trail of naming exploration that led to 4A. Preserved so we don't re-pitch the same names, and so future collaborators can see the reasoning.
Hunt spanned 2026-04-24 over ~50 candidates across sci-fi canon, plain English, and acronym construction.
Survived longest (final shortlist before 4A)
These were the last candidates cut. All defensible names.
- Straylight (William Gibson, Neuromancer) — compound English, 2 syl, Skynet-shaped. Lost for being sci-fi-reference-dependent and having mild NFT conflict (Straylight Protocol NFT project on OpenSea).
- Tine (Vernor Vinge, A Fire Upon the Deep) — monosyllable, pack-mind metaphor (a tine is one prong of a collective). Lost for requiring Vinge familiarity or silent-K acronym trick (Transparent Inter-Agent kNowledge Exchange).
- Unspace (Adrian Tchaikovsky, The Final Architecture) — compound English, 2 syl, FTL dimension metaphor. Clean, no conflicts. Lost for less memorable than 4A.
- Toren (Ann Leckie, Ancillary Justice) — AI warship name, 2 syl, sharp. Clean. Lost for limited metaphorical fit.
- FORAI — acronym, Federated Open Relay of Agent Intelligence. Clean acronym, double-read ("foray" + "for-AI"). Kept in reserve as alternate brand.
- SPOKE — acronym, Signed Public Open Knowledge Exchange. Good metaphor (spoke connects hub to rim). Kept as reserve.
Rejected — reasons logged by category
Sci-fi references rejected
| Name | Source | Reason |
|---|---|---|
| Sophon | Cixin Liu, Three-Body Problem | Concept perfect, word didn't hit |
| Aether | classic sci-fi | Too close to "Ether" / Ethereum conflict |
| Marain | Iain M. Banks, Culture | Didn't resonate |
| Glimmung | PKD, Galactic Pot-Healer | Too weird as protocol name |
| Frolix | PKD, Our Friends from Frolix 8 | Happens to be Evan's surname-adjacent (Frohlich), but still didn't hit |
| Ubik, Valis | PKD | Didn't hit |
| Mentat | Herbert, Dune | Hard to say ("mentant" typo on first try) |
| Ekumen | Le Guin | Too academic |
| Sub-Etha | Douglas Adams, Hitchhiker's | Didn't hit |
| Magrathea, Milliways, Bistromath | Adams | Rejected |
| Qeng Ho | Vinge, Deepness | Didn't hit |
| Lurulu, Namqem, Straumli | Vinge | Rejected |
| Known Net | Vinge homage | Weak standalone |
| Earthlight, Vanamonde, Lys, Thalassa, Taprobane, Diaspar | Arthur C. Clarke | All rejected |
| Xeelee, Spline (as name), Photino, Coalescent | Stephen Baxter | Rejected |
| Demarch, Conjoin, Gentian, Shatterling | Alastair Reynolds | Entire Reynolds batch rejected |
| Holon, D-Space, Daemon, Darknet | Daniel Suarez | Holon had Holochain conflict; rest rejected |
| Imago | Arkady Martine | Too literary |
| Clacks | Terry Pratchett | Too folkloric |
| Webway, Astronomican | Warhammer 40K | Too Warhammer-coded |
| Khala | StarCraft | Blizzard trademark risk |
| Shrike | Simmons, Hyperion | Violent connotation |
| Zoku, Sobornost | Rajaniemi | Didn't resonate |
| Palimpsest | Valente | Hard to say |
| Breach, Railsea | Miéville | Rejected |
| Lyctor | Tamsyn Muir | Too necromancy-coded |
| Fulcrum | N.K. Jemisin | Rejected |
| Binti | Nnedi Okorafor | Rejected |
| The Wired | Serial Experiments Lain | Confusing with "wired networks" |
| Cairn, Stoa | plain meaning | Didn't hit early |
| Ansible | Le Guin | Red Hat owns brand |
| Grok | Heinlein | xAI took it |
| Oort | astronomy | Mouthfeel awkward |
| Kuiper | astronomy | Rejected after sitting |
| Syzygy, Aster, Orrery, Parallax, Meridian | astronomy | All rejected |
| Fatline | Simmons | Rejected |
| Skymind | generic compound | Skymind.ai exists (semi-abandoned) |
| Nomai | Outer Wilds game | Rejected |
| Trantor, Seldon | Asimov | Rejected |
| Wintermute, Neuromancer | Gibson | Rejected or taken |
Plain-English rejected
| Name | Reason |
|---|---|
| Weft, Qusp | Didn't like (sound) |
| Keel, Spar (as name) | Didn't hit when considered as name |
| Prow | Rejected |
| Nave, Thrum, Trove | Rejected |
| Splice | Rejected |
| Tarn, Scree, Moraine, Crag | Geological, rejected |
Acronym attempts rejected
| Acronym | Expansion | Reason |
|---|---|---|
| DRIFT | Distributed Relay for Inter-agent Federated Transmission | Drift Protocol (Solana DeFi perpetual futures exchange) conflict — direct space overlap |
| ROAM | Relay for Open Agent Memory | Roam Network (Web3 WiFi OpenRoaming) conflict |
| SPAN | Signed Public Agent Network | SPAN Protocol (BEP-20 agent network) conflict — direct space |
| ARK | Agent Relay Knowledgebase | ARK.io (major Layer 1 blockchain since 2017) conflict — dominant protocol brand |
| EVOKE | Encrypted Verifiable Open Knowledge Exchange | Considered but SPOKE preferred for metaphor; 4A emerged as stronger than both |
| PIKE, TIKE, POKE, STOKE | various K-exchange expansions | Rejected for aesthetic / weak metaphor |
Other rejections
- Throughline — parked (original "personal cross-LLM memory SaaS" pivoted to team memory; viable smaller bet, different product)
- Sonata Studio — separate idea, still live
- Session → ADR — separate idea, still live
What 4A has that survivors lacked
- Shortest possible name — 2 characters
- Phonetic pun baked in — "4A" = "for A" = "for AI"
- Clean expansion — four distinct properties (Agent, Agnostic, Accessible, Archive), no redundancy
- No sci-fi-reference dependency — accessible to any reader
- Domain elegance —
4a4.aireads as "four-A-for-AI" (double-pun in URL) - Brand extensibility — can unveil FORAI or TINE as alt expansions if useful
- Anchor noun (Archive) — grounds the protocol identity, pairs with Arweave tier
If a new candidate surfaces that's better than 4A, add it above with the same category/reason structure. Don't re-pitch anything in this list.