header image

Manifold Community Fund

0 proposals
0 active projects
$14K

You can see the original Notion doc here, which includes explainer images and allows you to leave comments.

Overview

Manifold is giving out up to $30k in prizes for community work done over the next 3 months. And you can invest $ or mana into the success of these projects, via impact certs!

How does an impact cert work?

  1. A founder posts a project proposal on Manifund, with the minimum funding they need to work on the project.

  2. Investors can read the project proposal, and offer the founder money in exchange for equity (a share of the prize payouts).

  3. If the project hits its minimum funding goal, it becomes active: investors receive equity, the founder receives cash, and the project is now underway

    • If a project does not reach the minimum funding goal, but the founder ships the project anyway, they will still be eligible for prize payouts.

  4. Throughout the next three months, the project founder posts updates on their progress, allowing investors and the Manifold team to evaluate their work and trade accordingly.

  5. Once per month, around the payout dates (Dec 15, Jan 15, & Feb 15), the Manifold team will evaluate all projects and award prizes.

For more background on impact certs, see Manifund’s explainer and the last ACX round.

Goals

  • Reward sustained contributions by community folks

  • Encourage new, high-quality projects to legitimize PMs

  • Discover grassroots ways to promote and grow Manifold

  • Find people we may want to hire in the future

  • Test out impact certs & set up a robust cert ecosystem

Prize payouts

  • The Manifold team will be deciding the prize amounts based on perceived value to the Manifold company

    • Exact methodology TBD; probably using the Manifund retro eval algorithm we’ve developed, which weighs inside-view estimate & confidence

    • But could be quadratic vote or median of evals (like in ACX Minigrants)

  • Ideas are cheap, execution is everything!

    • We’ll award prizes based on the quality of the output, not on how good the idea was.

  • Each payout aims to assess impact realized between each checkpoint

    • For example: if someone starts a great podcast on Jan 1 2024, they might earn $1k for 2 episodes before Jan 15, and then $2k for 4 new episodes before Feb 15.

    • Existing projects are eligible — you can create impact certs for these and offer to sell them!

  • The prize will be in the form of a buy offer on certificates, which sets a price floor. Investors and founders can sell against this floor, or wait for a better offer in a future round.

  • Payouts take the form of:

    • Cash for project founders, which can be withdrawn

    • Charity balance for investors, which can be donated or further invested

  • We reserve the right to update prizes and structure of the fund as necessary

    • For example, if we don’t see enough good projects creating impact early on in Nov, we might distribute the prizes as $5k/$10k/$15k across the 3 checkpoints

Rough heuristics for prize payouts

(Warning: very fuzzy, these are not authoritative or guaranteed)

  • Manifold would be happy to pay $2 per real user signup

    • (with perhaps some threshold for transaction costs, eg minimum 25 new users)

    • Or $50 per notable user (eg 100k+ Twitter followers) who creates & shares a market, or someone who is Trustworthy-ish quality

  • $300 for a high-quality blog post promoting Manifold

    • eg one that gets cited by some reputable external source, like the WvM writeup

    • (we’re generally very open to guest posts)

  • $100 for software eng work that we like and would have taken a core team member 1 hour

    • (warning: due to onboarding and switching costs, a good SWE might expect to take 2-10x as long as one of our own team members)

  • Don’t Goodhart on these too much, though!

    • Austin thinks the best projects come from people who want to do something awesome, and we’ll reward projects even if they don’t fit in a neat bucket

    • Feel free to ask about other kinds of metrics or evaluations

FAQ

  • Q: As a creator, how much effort should I expect to put into my project?

    • Expect to put in 10h+ for a particular project!

  • Q: What’s the relation between “$”, “mana” and “Manifund balance”?

  • Q: Should someone with lots of projects create a single entry for themselves? or per project?

    • Lean towards one “umbrella project” for you, instead of many mini projects. (We also encourage you to form teams!)

  • Q: Do I have to offer to sell some of my equity to be eligible for the prize?

    • Yes, at least enough to seed the impact cert AMM on Manifund. Part of the goal of the round is to test impact certs.

    • Projects that didn’t hit their minimum funding bar on Manifund, but actually executed, will still be eligible for prizes.

  • Q: Does my project have to pertain specifically to Manifold?

    • Not necessarily. For example, a widely-read blog post about using prediction markets within your startup would be eligible for funding, if we think the knock-off effects on the forecasting ecosystem and thus Manifold are good.

  • Q: If I’m not sure about how impact certs work, how can I test them out?

    • Go to dev.manifund.org and create your project there! Austin can send you some fake money to use on dev to test things out.

Project ideas

Here’s a flavor of the kinds of projects we’d be excited to see!

  • Organize & moderate debates that incorporate live-betting on markets

  • Host a weekly call on Discord to welcome new Manifold users and help them get oriented

  • An Actually Good podcast series interviewing top users and team members

  • Run a Manifold game night at your local rationality meetup and posting a writeup (like this)

  • Build a good matching algorithm for pairs of profiles on manifold.love

  • Host a premier prediction market tournament on the Israel vs Hamas conflict

  • Publish a daily short Substack on interesting happenings in the world of forecasting (cf Matt Levine)

  • Set up an unofficial Manifold swag store using printful and DallE art

  • Produce 5 short YouTube videos on “Getting Started on Manifold”

  • Creating a popular website or mobile app that wraps Manifold, using our API

  • Incorporating prediction markets into a livestream of chess gameplay on Twitch or Discord

  • Analyze the forecasting track record of various prediction market platforms, polls, pundits, etc.

  • Build an OpenAI assistant for Manifold

  • Market Manifold or Manifold.love online or in person and get many users to signup

Examples of great community projects

  • Major contributions to Manifold’s growth:

    • Scott Alexander’s “Mantic Mondays”

    • QuantumObserver’s LK99 market

    • Memestiny’s Destiny Orbiter stocks

  • Mira’s GPT4 market: This was a cool way to showcase prediction markets as a crowdsourced challenge. Mira put in a lot of effort into setting up good criteria and verifying the outputs, and the market got super popular (hitting #1 on Hacker News)

  • JSKF’s mirror bot: Lots of people have suggested mirroring more markets between Manifold and other prediction sites, but JSKF sat down and figured out all the logistics in doing so, just by using our managram API in a clever way.

  • Isaac King’s search dashboard: Our search functions can be somewhat lacking at helping you find exactly that market that you were looking for. Isaac saw this problem, and built out a third-party search utility using our API, which gets regularly used by Manifold community members.

  • WasabiPesto’s Calibration City: A big bottleneck to wider adoption of prediction markets is, “can we trust these things at all?” WasabiPesto built a very configurable site to answer that very question, letting people understand when prediction markets are calibrated.

  • Tobias’s playing cards: A beautiful example of custom art made for the Manifold community, which draws on the lore of prediction markets, stock trading, and forecasting and turns them into cute on-brand animals. We relied on these as the awesomest swag we had at the Manifest conference!

  • Marketwise blog: Marketwise explores a variety of topics that are interesting to the Manifold community, on a regular cadence. One of his posts, on the benefits of boosting vs subsidy, was well-researched and served as the impetus for Stephen to simplify the boosting system!

  • Market Manipulation Podcast: This was originally a joke by Conflux, but got more and more elaborate over time… until it morphed into a serious effort at producing good episodes! And bonus points for Conflux’s in-depth writeup of their production efforts, posted on the official Manifold newsletter~

  • Eliza’s moderation practices: Eliza is our most active moderator by far, and has spent countless hours resolving minutiae on so many markets and weighing in on Discord. (Sometimes I wonder if he’s secretly a front for an entire team of users. Or a robot.)

  • Jack’s 2022 midterms prediction market performance analysis: The accuracy of prediction markets vs prediction polls is a longstanding one, and Jack attacked this question directly using on a highly-traded and socially valuable dataset.

  • See also the Community Bounties we’ve historically paid out!

Feedback

If you have questions or thoughts, come share them on the Manifund Discord!