For the people, by the people

Turn surveillance around.

The state and private surveillance companies can assemble years of observations into a case. Ordinary people rarely have an equivalent historical record. The answer is not another centralized surveillance database. It is an open, decentralized evidence protocol built for evidentiary symmetry.

The state says: “Our records show what happened.”

Society should be able to answer: “Good. Let’s compare them with ours.”

Screenshot discussing plans to turn rideshare and delivery vehicles into mobile license plate readers
The same distributed sensor idea can produce very different power depending on who owns the data and what the system is designed to answer.
01 / The problem

Surveillance is asymmetric.

Institutions accumulate observations by default. Citizens usually encounter evidence only after an accusation, a collision, a protest, an arrest, a lawsuit or another disputed event. By then, the useful camera may be unknown, overwritten or controlled by the same institution whose account is being challenged.

Institutions remember

ALPR hits, bodycams, incident reports, databases, CCTV and private vendors can be combined into a timeline.

People must already know

Discovery is weak when you do not know which camera, vehicle, database or recording existed in the first place.

Aggregation creates the story

Each observation may be trivial. Once linked by time, place and identity, they become a narrative with institutional weight.

02 / The inversion

Use the same sensor density. Change the direction of power.

Centralized surveillance
Public movement → private sensors → centralized database
→
Primary question
“Where has this person been?”
Public witness network
Public movement → distributed sensors → open protocol
→
Primary question
“What independent evidence exists about what happened here?”

The distinction is between tracking people and preserving independent witness.

03 / The protocol

Do not build a master database. Build a common language.

Cars, buses, bicycles, taxis, delivery vehicles, phones and storefront cameras already generate observations. The protocol allows independent devices and storage operators to prove that evidence existed at a particular place and time without automatically making raw footage publicly searchable.

OBSERVATION { timestamp: 2026-08-16T20:43:17Z coarse_location: intersection-or-zone source_type: vehicle_camera evidence_hash: 7f3a...9c21 storage_replicas: 3 public_index: metadata_only raw_media: encrypted }
1

Capture

A participating device records an observation as it normally would.

2

Commit

The device creates a cryptographic fingerprint and timestamp so later alteration can be detected.

3

Replicate

Encrypted evidence is stored across several unrelated nodes instead of a single institutional archive.

4

Index existence

The public can discover that observations exist for a place and time without automatically seeing their contents.

5

Disclose narrowly

Relevant recordings can be released through consent, legal process, journalistic/public-interest rules or another transparent governance mechanism.

6

Verify

Anyone receiving the evidence can check that the disclosed file matches the fingerprint committed at the time of recording.

04 / What discovery becomes

You should be able to discover evidence you did not know existed.

QUERY place: Broad St & 11th Ave time: 20:35–20:50 RESULT 27 independent observations preserved 9 vehicles 3 bicycles 2 businesses 1 bus 12 other sensors RAW CONTENT Encrypted / restricted NEXT Request disclosure for the specific event window

The network reveals the existence of independent observations first. That changes the discovery problem. A person no longer has to know that “Camera 927” existed before they can establish that evidence may contradict an institutional account.

05 / What it can challenge

A distributed historical record changes who can reconstruct an event.

Police encounters

Compare official reports and bodycam footage with unrelated observations captured before anyone knew they would matter.

Protests

Reconstruct sequence, timing and movement from many independent witnesses instead of one institutional camera system.

Traffic collisions

Resolve disputed lights, movement and timing using unrelated cameras around the intersection.

Journalism

Find preserved evidence months later without depending entirely on police, corporate PR or a lucky eyewitness.

Civil disputes

Verify whether vehicles, workers, contractors or other actors were present at a claimed time and location.

Public accountability

Give communities a record that does not disappear simply because an institution chose not to preserve its own footage.

06 / Safeguards

Do not democratize stalking.

A public evidence commons fails if it becomes “Flock for everyone.” The protocol must be intentionally bad at constructing persistent dossiers about individuals and intentionally good at verifying bounded events.

Index place + time, not person + history.Search should begin from an event window, not a person, plate or identity dossier.
Encrypt raw media by default.Public discoverability should expose the existence of evidence, not unrestricted access to everyone's movements.
No central operator.Independent nodes prevent the institution being challenged from controlling the only copy.
Open source everything.Capture software, timestamping, retention rules, query logic and verification tools must be inspectable.
Make tampering evident.Cryptographic fingerprints allow later recipients to determine whether evidence was modified after capture.
Use bounded retention.Retention should be governed by transparent rules, with longer preservation only when an event is specifically flagged.
Design test: If the system is excellent at answering “Where did this person go all week?”, it has reproduced the surveillance architecture. If it is excellent at answering “What independent evidence exists about this disputed event?”, it is moving toward evidentiary symmetry.
07 / Governance

The archive should have no single owner.

Storage and verification could be operated by a federation of independent actors rather than one government agency or surveillance vendor.

civil-liberties groups universities journalists co-ops unions municipal nodes community organizations individual storage nodes

For centuries, powerful institutions possessed archives while ordinary people possessed memories. Digital technology can either deepen that imbalance or help reverse it.

08 / The proposition

Nobody gets a monopoly on the historical record.

Surveillance power over

  • Centralize observations
  • Resolve identity
  • Build persistent histories
  • Restrict access
  • Let the database define the narrative

Public witness power to verify

  • Distribute custody
  • Preserve independent observations
  • Search bounded events
  • Reveal existence before disclosure
  • Compare institutional claims with outside evidence
Pass the witness

Share the protocol.

Choose a tile. Each opens a ready-to-post message on the network you use.

0:00