An institute for the questions that still resist us.
Lightcone Institute is the Foundation's program layer: research collaborations, prize problems, review panels, and AI-enabled scientific workflows, with an initial research spine in gravitational-wave physics, effective field theory, and generative models for scientific discovery.
A living prize commons for open problems.
Horizon is a public challenge system for problems important enough to deserve money, attention, and independent verification.
This is not a vertical canon of problems chosen once by one institute. It is closer to a research frontier with a public memory: scientists, engineers, and theorists can propose the questions they believe matter most.
Lightcone curates accepted problems, publishes clear statements and solution criteria, promotes them to the research community, and helps attach prize pools to the questions donors want solved.
Researchers submit frontier problems
Each proposal explains the question, why it matters, what counts as progress, and what would count as a solution.
Lightcone shapes a public dossier
Accepted problems receive a statement, field tags, references, review criteria, and a transparent prize status.
Donors choose exact prize pools
When the donation system is live, donors will be able to direct gifts or pledges toward specific accepted problems.
Solutions face expert review
Claims are checked by a review panel. Prizes are awarded only after the stated standard has been met.
Donor interest should follow curiosity.
Some donors care most about physics. Others care about AI verification, quantum algorithms, autonomous laboratories, or tools that make scientific work faster and more reliable. Horizon lets that preference become visible.
The public page for each accepted problem will show its statement, proposer, prize pool, review standard, and current status. The goal is not only to award prizes, but to focus attention where attention can still change the future of science.
Problem statement
Question, field, references, solution criteria
Prize pool
Foundation seed funds plus donor-directed gifts
Review panel
Domain experts selected for each accepted challenge
Progress record
Submissions, milestones, awards, public updates
Plural by design
Horizon can host problems from many proposers instead of preserving one fixed institutional list.
Clear standards
Every prize problem needs a published bar for success before money or attention starts to gather around it.
Open science first
Prizes should create public knowledge, not private trophies. Results and explanations should be accessible.
Concrete programs at the physics-AI boundary.
Lightcone's initial research directions grow out of work on effective field theory for gravitational-wave physics and AI methods for waveform generation, detector analysis, and scientific representation learning.
The operating model is built for grant-backed research partnerships: focused scientific challenges, reproducible benchmarks, data and simulation workflows, independent validation, and open outputs.
EFT for compact binaries
Finite-size effects, spin, dissipation, parity-violating signatures, and waveform corrections for systems where subtle physics can become observable.
Compact objects in media
Point-particle and fluid EFT methods for modeling binaries moving through matter, viscosity, and other environmental effects.
Adaptive searches
Transfer-learning workflows for gravitational-wave data that can adapt to changing, non-Gaussian, and time-dependent detector noise.
Waveform and latent-space discovery
Generative models for faster waveform production, richer template banks, and dual-space representations that expose useful scientific structure.
Bring a problem worth solving.
Researchers can send problem proposals, possible review criteria, or collaboration ideas. Donors and partners can discuss how to support specific scientific challenges as the prize infrastructure comes online.