Humanity has a backlog

Useful work
still exists.

Important problems are full of unfinished work. yah.quest is building a way to make that work visible, bounded, verifiable and honest about compensation.

Checking current opportunities… The scout is loading its latest mirror. External opportunities and yah.quest-owned work packages are tracked separately.

Start with real need, not a job title.

Explore where human judgment, care, presence and accountability are still useful, then preview how matching could work.

Bring one bottleneck a real team owns.

yah.quest is looking for bounded public-purpose work that an organization can safely externalize, verify and use downstream.

Problems indexed100
Prototype questlines3
Prototype work packages28
Open external opportunities…

Cancer

A huge problem is not a task. The prototype shows how cancer-related work could be broken into bounded, non-clinical contributions with explicit safety and verification rules.

Vaccine development

Vaccine R&D is not only bench science. Public roadmaps, trial registries, manufacturing readiness and access constraints contain bounded coordination work that can be done safely under expert review.

Make personalized medicine available to everyone

Personalization can reach molecular precision while access remains painfully uneven. This prototype maps the infrastructure around precision medicine: access, individualized manufacturing, genomic-data governance and affordability.

Find work

Where are humans
still needed?

AI changes which tasks are cheap. It does not erase the backlog. Start with places where another human hour can create useful marginal value.

Checking current external opportunities…

The Opportunity Scout is loading its latest primary-source mirror.

Loading current work

External leads

Loading the latest synchronized opportunity mirror…

01

Judgment

Evidence can be generated quickly. Someone still has to choose what matters, what is safe and what happens next.

02

Care

Health, education and public services involve people, trust and consequences, not just outputs.

03

Presence

Land, infrastructure, ecosystems and communities contain work that cannot be completed from a data center.

04

Accountability

Useful work needs someone willing to own the result, review it, integrate it and answer for what happens next.

What could you contribute?

Tell the matcher your skills, time, location and compensation requirement. It explains why a prototype surfaced.

Problems

Explore the
backlog.

The 100 issues are an index, not an objective moral ranking. Most do not yet have owned, verifiable work packages.

How much does this matter?

Future evidence can include burden, urgency, neglectedness, tractability and equity. Unsupported scores stay blank.

Can another human hour help now?

A problem becomes actionable only when there is ownership, safe decomposition, verification capacity and a clear compensation status.

Area
View

Most problems are still indexed only. Open a prototype questline to see what safe decomposition looks like.

Cancer prototype
001

Health / Cancer

One problem.
Many bounded tasks.

The metaphor stops where safety starts.

This is a research-grounded prototype, not an endorsed or clinical workflow.

Scientific Reports ↗ NCI ↗ Cancer Grand Challenges ↗ WHO / TDR ↗
Mission ownerMissing
CompensationNot decided
Review capacityNot staffed
Safety boundaryPublic / cleared data only
Human-time bottlenecks are real.

Peer-reviewed cancer research has shown that trained non-experts can produce useful outputs in a tightly bounded workflow. In one 2026 breast-cancer pathology study, 280 participants produced 162,020 annotations using repeated labels and strong ground truth.

Demand for a yah.quest crowd is still a hypothesis.

We have not proved that cancer labs want outside contributors, that coordination overhead is worth it, or which backlog matters most. The work packages below are prototypes, not requests from researchers.

Find one lab with a real backlog.

Before recruiting contributors, ask: “What useful task would you do if someone gave you 1,000 careful human hours?” A backlog only becomes live when the owner confirms it is genuinely deferred by limited capacity, safe to externalize, objectively verifiable, supported by reviewer capacity, clear about compensation, and connected to a real downstream workflow.

Reduce cancer deaths and suffering

Earlier detection

Non-clinical phase-one boundary

No diagnosis, personalized treatment, unsupervised trial-eligibility decisions, identifiable patient data or open biomedical experimentation.

Vaccine development prototype
002

Health / Vaccine development

Faster vaccines.
Better coordination.

Useful work exists outside the lab.

This track turns public R&D priorities, evidence, readiness and access bottlenecks into bounded work under expert review.

WHO vaccine R&D roadmaps ↗ WHO PPCs + TPPs ↗ CEPI 100 Days Mission ↗ WHO vaccine pipeline ↗
Mission ownerMissing
CompensationNot decided
Review capacityNot staffed
Safety boundaryPublic, non-operational work
Accelerate safe, effective and accessible vaccines

R&D priorities

Non-lab phase-one boundary

No pathogen handling, genetic engineering, experimental vaccine design, dosing or formulation instructions, trial-eligibility decisions, identifiable participant data or unsupervised biomedical experimentation.

Personalized medicine prototype
003

Health / Precision medicine access

Make personalized medicine
available to everyone.

Molecular personalization is not social access.

This research-grounded prototype asks what has to be built around the science so biological uniqueness does not become a new layer of inequality.

WHO precision medicine resolution ↗ WHO genomics equity analysis ↗ WHO genome-data governance ↗ NCI precision medicine ↗ Moderna individualized therapy case ↗
Mission ownerMissing
CompensationNot decided
Review capacityNot staffed
Safety boundaryPublic, non-clinical work only
Mass-produce uniqueness without mass-producing inequality

Access infrastructure

Infrastructure-only phase-one boundary

No diagnosis, therapy selection, interpretation of individual genomic results, trial matching, identifiable patient or genomic data, wet-lab procedures, experimental molecular designs, dosing or formulation instructions.

Prototype matcher

Prototype matcher

What could
you do?

This is decision support, not an invisible boss. It explains recommendations and does not turn a score into a judgment of your worth or pay.

Prototype matcher only.

The external scout leads on Find work are real open calls, but this matcher still ranks only prototype work packages. Nothing here submits an application or reserves work.

1. What can you already bring?
2. What languages can you work in?
3. How much time do you actually have?
4. Where can the work happen?
5. What kind of work?
6. What do you need from compensation?

How it works

Useful work needs
an owner.

The product is not a giant task marketplace. The atomic unit is a work package that somebody genuinely needs, can review and can use.

01Problem

A public-purpose need worth addressing.

02Owner

A real organization owns the bottleneck.

03Work package

Output, scope and acceptance criteria are explicit.

04Person

Matching is explainable and optional.

05Verification

Review capacity is real labor and is budgeted.

06Compensation

Paid, volunteer, stipend or prize is stated honestly.

07Integration

The accepted output enters a real workflow.

08Evidence

We track what happened downstream.

Do bounded work.

Know the output, expected effort, compensation status, reviewer and acceptance criteria before you start.

Bring a real bottleneck.

Define work that your organization can safely externalize, review and actually use.

Review is work.

Apply a declared protocol, record uncertainty, request revisions and escalate ambiguous cases.

Fund the whole chain.

Contributor and verifier funds should exist before paid work opens, not after the labor is done.

Quest sheet

Work-package protocol

C-ED-01

Work package

Provenance
Status
Owner
Expected output
Work type
AI assist
    Protocol
    Reviewer
    Capacity
    Review time
    Escalation
    Evidence model

    Evidence

    Submission is
    not impact.

    Keep the states separate so activity cannot masquerade as usefulness.

    1Submitted

    An artifact exists.

    2Verified

    It passed its quality protocol.

    3Accepted

    The owner accepted it.

    4Integrated

    It entered a real workflow.

    5Used downstream

    Someone used it later.

    6Outcome evidence

    A later public outcome is observed.

    Prototype evidence rows only.

    These demonstrate the model. They are not completed yah.quest contributions.

    Research

    Evidence before
    metaphor.

    Citizen science, humanitarian mapping, public employment, platform work and older project history all inform the operating model. None proves it works yet.

    The missing layer is not recruitment alone. It is connecting ownership, decomposition, verification, compensation, worker rights and downstream evidence.

    The coordination problem predates yah.quest.

    Archive material reaches back to earlier work on youth, public problems, cleanup coordination and new funding rails. Those projects were not earlier versions of yah.quest, but they repeatedly exposed parts of the same missing middle.

    Read the source-backed origin note ↗

    The repo keeps the receipts.

    Research notes record adjacent systems, labor-transition evidence, cancer safety, vaccine-development decomposition, personalized-medicine access, funding and governance, partner discovery and source handling.