Hackathon Pack 2026
Three problems ready to hand out as challenge statements: sourced, scoped, and briefable in a paragraph.
3 problems
Skip to content
The IndexCollections
Collections
Named sets of problems: packs chosen by hand, and every domain, theme and horizon the Radar files against.
Picked by hand
When the machine decides
What happens when software acts at scale and the checks around it still run at human speed.
3 problems
The measurement gap
Problems where the fix is not invention but instrumentation: something important is happening and nobody counts it where it counts.
4 problems
By domain
Energy
Problems in power, grids and the energy transition worth taking on.
3 problems
Health
Problems in care, access and medical administration worth solving.
2 problems
Finance
Problems in credit, payments and fraud that reach small firms and households.
2 problems
Work
Problems in hiring, skills and how work is changing under AI.
2 problems
Cities
Problems in housing, heat, infrastructure and urban life.
2 problems
Enterprise AI
Problems created as AI and automation arrive in real companies, from delegated decisions to automated attacks.
2 problems
Logistics
Problems in trade, shipping, returns and the physical economy.
2 problems
Education
Problems in schools, learning and what AI did to both.
1 problem
Public sector
Problems in government services, procurement and civic capacity.
2 problems
Media
Problems of provenance, attention and what is real online.
1 problem
Food
Problems in farming, supply and what ends up on a plate.
1 problem
By what is broken
Coordination
Problems where the pieces exist but nothing joins them up.
2 problems
Measurement
Problems where something important is happening and nobody counts it.
8 problems
Access
Problems where a service exists but the people who need it cannot reach it.
4 problems
Trust
Problems where you cannot tell whether to believe what you are looking at.
2 problems
Capacity
Problems where the need is obvious and nobody has the hands to meet it.
3 problems
Recourse
Problems where a decision is automatic and the appeal is manual.
1 problem
By what it would take
A hackathon
Problems where a weekend and a few people could put something real in front of a user.
5 problems
A side project
Problems a small team could carry through evenings and weekends until they work.
6 problems
A funded startup
Problems that need a company behind them: capital, a sales motion, and the right to operate.
9 problems
A moonshot
Problems no single company can finish, because they need capital and a coalition.
0 problems