What it is
Evidence capture
A GPS-pinned, timestamped submission path for screenshots, footage, vehicle context, environmental notes, and observer statements.
Elon-facing first-pass pitch package
Machine perception keeps seeing the edge of the map.
Phantom Signal is a public evidence-and-review layer for moments when Tesla, EV, or ADAS perception systems appear to render or classify human-like figures or anomalies where human observers report no visible cause.
Not proof. A signal.
Live now: this Elon-ready pitch experience is deployed at phantom-signal-elon-pitch.pages.dev. The older project-packet site and broken custom domain status are documented near the bottom.
The machine perception gap
ADAS systems do not “see” like humans. That gap is the story.
Modern vehicles combine camera, radar, ultrasonic, GPS, and model inference. In unusual lighting, reflective surfaces, occlusion, weather, graveyards, roadsides, bridges, and empty lots, a system may appear to detect a person-shaped object that no witness can confirm. Phantom Signal turns those moments into structured, reviewable cases.
What it is not
No ghost proof claims
The default posture is “likely explained.” Language stays inside machine-perception anomaly, probabilistic inference, and sensor-fusion review.
Why now
Public curiosity meets autonomy
As vehicle AI becomes a daily interface, unexplained detections become a powerful way to educate, audit, and engage the public around perception limits.
Demo journey
A turnkey public experience that can be sent as a single link.
The first pass should feel like a cinematic case-room, not a claim machine: open with the thesis, show sample signal cards, invite structured submissions, and route serious partners to the deck/operator workflow.
Land
Visitor sees “Not proof. A signal.” and understands this is independent, cautious, and evidence-first.
Explore signal cards
Map/list cards show anonymized examples: location type, vehicle context, media type, confidence language, review status.
Submit or join
Scouts upload sightings; Owners offer vehicles/locations/footage; Operators request review or licensing workflows.
Review protocol
Every case passes through safety, consent, data hygiene, and explanation checks before publication.
Escalate
Promising cases become field missions, partner demos, short-form episodes, or research collaborations.
Roles
Scout, Owner, Operator: three ways into the network.
Scouts
Submit GPS-pinned media evidence and witness notes from real-world vehicle perception moments.
- Capture screen/video safely while parked or as passenger.
- Record time, weather, direction, vehicle mode if known.
- Accept that cases may be explained or rejected.
Owners
Contribute vehicle, location, access, dash footage, or private-lot testing environments.
- Support controlled repeats without unsafe driving.
- Confirm permissions for location and media use.
- Help compare human observation vs vehicle rendering.
Operators
Review, host, license, or produce future Phantom Signal experiences.
- Moderate submissions and field protocol.
- Package cases into demos, episodes, or partner briefings.
- Maintain legal, safety, and non-affiliation boundaries.
Field protocol
Capture the anomaly without creating a new one.
- Safety first: no distracted driving, trespassing, harassment, road obstruction, or late-night risky field work.
- Minimum packet: GPS/location type, timestamp, media, observer note, vehicle/ADAS context if available, and environmental conditions.
- Explain first: look for people, signs, reflections, shadows, animals, UI artifacts, map data, roadside objects, and weather effects.
- Preserve originals: keep raw screenshots/video and avoid edits beyond redaction.
- Respect privacy: blur faces, plates, home addresses, minors, and sensitive property details.
- Review status: label as likely explained, under review, unresolved, or rejected — never “proven.”
Safety + boundary
Independent and non-affiliated
Phantom Signal is an independent concept and is not affiliated with, endorsed by, sponsored by, or officially connected to Tesla, xAI, Elon Musk, or any automaker. Brand names may be referenced only descriptively when discussing user-submitted vehicle/ADAS observations.
No official logos, UI mimicry, private data scraping, endorsement implication, or “proof of ghosts” language belongs in the public experience.
Elon-facing ask
Attention, feedback, and a path to test the edge cases — not an endorsement request.
1. Look
Review the concept: a public, cautious archive of machine-perception anomalies that people already screenshot and debate.
2. Pressure-test
Tell us what would make the protocol useful, credible, or immediately disqualifying from an autonomy/perception perspective.
3. Collaborate
If there is interest, support a controlled pilot with data hygiene, vehicle access, or expert review — no endorsement required.
10-slide deck outline
The deck should sell curiosity with restraint.
A concise deck can live in-page first, then become PDF/keynote after the demo path is live.
Next-build checklist
- Replace current Pages project-packet page with Elon-ready landing experience after approval.
- Fix custom domain Cloudflare 530 / 1033 before using
ps.breedloveforever.comin outreach. - Build demo map and signal-card data schema with sample-safe placeholder cases.
- Add submit/join forms for Scout, Owner, Operator interest with consent language.
- Create deck request workflow and export the 10-slide outline to PDF.
- Draft moderation guide and legal/privacy review checklist.
- Connect newer MVP repo functions/D1 only after static pitch is approved.
Outreach link strategy
Use now for audit only: https://phantomsignal-site.pages.dev returns HTTP 200 but is not Elon-ready yet.
Do not send yet: https://ps.breedloveforever.com is currently broken with Cloudflare 530 / 1033.
Local package: this static artifact is designed to be reviewed locally, then copied into the live Pages project when approved.
IndependentEvidence-firstADAS anomaly reviewNo endorsement implied