Wyrd Labs Research programs · hypotheses · horizon Research is labelled research
Research

Wyrd Labs gives unconventional questions a serious engineering trial.

The divisions commercialize only the ideas that survive evidence. Everything on this page is labelled by what it currently is — an active program, a prototype, a hypothesis under test, or a recorded option.

The method

Innocence Translation.

It does not mean turning dangerous technology into a civilian product. It is a discovery method that protects curiosity without protecting weak ideas from evidence.

01
Put an intelligent person who has not been conditioned by the industry's assumptions in front of the problem.
02
Let that person ask questions experts may consider naive, impossible or already settled.
03
Translate the question into technical primitives.
04
Ask domain experts to attack it.
05
Build a cheap model, simulation, benchmark or prototype.
06
Compare it against strong baselines.
07
Check law, ethics, cost, safety and deployment reality.
08
Keep it only if the evidence survives.

Innocence asks the question. Expertise tries to kill it. Proof decides.

Programs

What is being tested, and at what class.

ICARUS Forge
Generate, test and compare governed agents or policies.
Research / integration
Question under test

Can agent and policy variants be generated, evaluated and selected under explicit criteria rather than hand-authored one at a time?

DAEDALUS Forge
Simulate plans, failures, counterfactuals and deployment blueprints.
Architecture / research
Question under test

Can rehearsal against a twin become a normal, cheap step between idea and pilot — instead of an expensive special project?

NINA Foundry
Generate and benchmark optimization methods under constraints.
Research prototype
Question under test

Does a portfolio of generated methods, verified by an assurance layer, beat a single brittle winner on real domain constraints?

Current honesty

The earlier framing that a language model simply "beats solvers" is retired. Prior measurement was found unreliable; the instrument is being repaired before any performance claim.

Substrate Hypothesis
Does nested constrained evolution produce useful reusable methods?
Scientific research
Status

A hypothesis, not a result. It is pre-registered internally and reported as unproven until an instrument we trust says otherwise.

Agentic genetics
Variation, selection, mutation and evaluation of agent strategies.
Research prototype
Status

Explored under controlled evaluation. The hypothesis is not proven.

Engine Corpus
Reusable library of solvers, classifiers, planners and specialist engines.
Platform direction
Rule of entry

Every engine needs an owner, a benchmark, tests and a consuming product. Otherwise it is not in the corpus.

Sovereign coding stack
Local or air-gapped coding agents with repo truth, proof and governed handoff.
Active R&D
What exists

Repo Truth Layer, coding proof packs, governed handoff between external agents, and time-boxed execution leases through PEAOS.

Proof systems
Receipts, ledgers, attestations, transparency, evidence packaging.
Active software
Direction

Signed evidence, hash chains, threshold controls, Merkle anchoring, and proof packs that a third party can check without trusting us.

OSS intake
Research, license, provenance, tailor, govern and integrate external open source.
Operating method
Why it is a program

Using open source responsibly is a licensing, provenance and maintenance commitment, not a download.

The horizon

What we explore.

Grouped by field. These are recorded directions — expansion optionality, not present commercial promises. Several will be killed by the Accuser's Office, and that is the point.

Sovereign intelligence
Cross-agent continuityLocal model packsSector skill packsFederation mesh4D access policy
Simulation & optimization
DAEDALUS CityDAEDALUS BuildDAEDALUS OpsScenario ForgeRed Team MirrorVessel routing
Cryptography & proof
Threshold signaturesHybrid post-quantum KEMMerkle transparencyPerfect-secrecy storageMPC on joint dataEthical black box
Human-machine systems
Wyrd LensSamael ExoDevice meshPrivacy wearablesHelmet & spacewalk support
Autonomous machines
HIVE degraded-mode coordinationPhaeton modular bodiesNOMAD persistent nodeCounter-autonomy verification
Health
Clinician councilsAceso clinical pathwaysBiota Vital closed loopsCrisis corridors
Cities
CTOS advisory pilotsInstitutional digital twinsAir readiness & U-spaceEmergency logistics
Oceans
Leviathan WatchHIVE DeepAcoustic & delayed syncCable and pipeline inspection
Space & Earth
Satellite proof packsARGOS Earth WatchInfrastructure drift monitorSpace Weather Sentinel
Sustainability
Proof of ComputeWorkload routingModel right-sizingHardware lifecycle
Two things this page is not

It is not a claim that any of these exist as products. And where a recorded idea has been parked or rejected — a from-scratch OS kernel, a phone manufactured as version one, a weaponized RADOME, an agent currency before authorization — it stays parked. The option portfolio demonstrates reach. It does not replace focus.

Restraint

Conflict as a quality mechanism.

An idea is not protected because it is ours. It is attacked, repaired, retested — and then it survives or it dies.

idea attack repair retest survive or die

The survivors go to the divisions. The rest are recorded and left alone.

If your institution has a hard problem and real constraints, that is exactly what the method needs.