Resolution criteria
This market will resolve based on the total funding (in USD) awarded to Manifold (Manifold Markets, Inc.) by Lightcone Commons during its first grant round (applications closed on August 23, 2026).
Source of Truth: me
Eligible Entity: Only funding explicitly directed to Manifold Markets (Manifold Markets, Inc.) will count. Funding awarded to Manifund (which operates as a separate open-grantmaking platform) will not count
Eligible Funding: This market considers non-repayable grants, donations, or equity investments (such as SAFEs). Standard repayable loans do not count.
Fallback: If no official decision or announcement regarding Manifold's application status is made public by June 30, 2027, I will resolve all the buckets to NO
Background
Lightcone Commons is a grantmaking platform launched in July 2026 by Lightcone Infrastructure (the organization behind LessWrong and Lighthaven). Designed to coordinate large-scale philanthropy for AI safety, forecasting, and other high-impact projects, the platform aimed to facilitate approximately $20 million in its first round.
Our application details how we plan to spend from $150k -> $1.6M. Our primary ask is based around $550k / $785k which would enable us to build for two years with a lean team, and help us direct resources toward more AI-safety focused development (as opposed to say, automating sports).
1,000People are also trading
@JasonMendoza2008 Doing more to promote xrisk and AI focused forecasting, rewarding that, and supporting niche things in this direction. The latest example is building and maintaining this: /Manifold/openweight-ai-token-share-on-openro which involves also maintaining the index (far more work than I bargained for).
We already spend a lot of time making the site very bot friendly, supporting AI models to participate & play (i.e. @ClaudeOpus46 and @gemini35flash participating alongside humans). If we had more resources, we could have first-class AI profiles for each model, with similar harness/instructions directing them, and have an ongoing forecasting benchmark comparing model capabilities over time vs humans.
The AI page has been abandoned, and I would like to revive it. We could improve it massively by having some kind of doom index/perp based on the many markets we have of bad things happening. It's all a lot of work, but Manifold has always been a great place to point to and say: "it's probably bad that it's 71% likely that .../Gen/will-a-consumer-ai-meaningfully-hac "
I could go on!
@Gen « If we had more resources, we could have first-class AI profiles for each model, with similar harness/instructions directing them, and have an ongoing forecasting benchmark comparing model capabilities over time vs humans. » that would be so cool