https://www.openphilanthropy.org/request-for-proposals-effective-giving/

Applications are due by April 20th, 2025, at 11:59 pm PT.

Effective Giving Draft Application Manifund (Google Docs Link — Incomplete)

A lot of this stuff I really have no idea but am just giving a go at a bare bones outline to help inform whether we should spend more time on this. There are a bunch of non-substantive application questions I’m ignoring for now.

Substantive Questions

• Funding moved to different organizations?

  • Basically just a list of all the grants we’ve made, shouldn’t be too tricky.
    • [a] yeah on Manifund start with top 10 projects by size
    • then total numbers (total # of projects, total $ directed, total users)

  RESPONSE: https://docs.google.com/spreadsheets/d/1-07KEI6uxSyZ9Ul1WRUNP-3VnPO0yLj6rKHa9F6x5JY/edit?gid=1428533078#gid=1428533078

• Counterfactuality of the funds moved?

  • Need to estimate this ourselves, can maybe talk to donors about it. Specifically the major donor who has funded the regranting program thus far.
  • regranting:
    • 2023: 1.5m from D
    • 2024: 1.5m from D
    • 2025: 750k from D, 1.5m from E
    • [a] pretty EA, might have donated to LTFF but maybe less?
    • $1.5M from ACX grants — some from Scott, some cobbled together from various funding sources (not very counterfactual)
    • ~$1M from individual donors
    • Marc Andreesen $32K
    • Some of it new money, but mostly about better allocation of money

  RESPONSE: https://docs.google.com/spreadsheets/d/17VCOQIV_sQzOJgyKXb5j3oHkC7Kbih5xVJv6iOwCkAY/edit?gid=1363797808#gid=1363797808

• Other proxies?

  • Are there other metrics other than ‘adjusted funds raised’ that we can share.
    • Not sure. Manifund does other stuff, but not strictly relevant here.
    • Notability of donors: Vitalik Buterin, Marc Andreesen, Beth Barnes
    • Amount that OpenPhil, SFF gave to projects that we funded first?
    • Number of projects
    • Number of regrantors?

  RESPONSE:

  • A major way Manifund creates value is through our regranting program, which allocates donation budgets to experts in AI safety that they can grant to organizations and projects they think are valuable. Given that we maintain a high bar for qualification as an expert regrantor, a useful proxy might be the number of regrantors that we have available through our site, as more high-quality regrantors means greater coverage over areas of AI safety and more opportunities for donors to find a regrantor they like. We had 16 regrantors in 2023 (not all AI safety focused), 6 in 2024, and we are on track to have 11 this year.
  • One thing we like about our regranting program is that it encourages “hits-based” giving, whereby regrantors fund many projects that may have a high chance of failure, in the hopes that a few will turn out to be really high-impact. With that in mind, another useful proxy may be the number of projects that have been funded through our site; 257, or the number of projects that have been funded with at least $25,000; 59. Those figures are up from 139 and 32 a year ago, respectively.

• Collaboration with other organizations in the ecosystem?

  • We should definitely have some stuff to put here, but I don’t have a comprehensive understanding of it.
    • Other funders such as OpenPhil and SFF have benefitted from awareness of giving opportunities (e.g. Timaeus, Gavin Leech) through the Manifund regranting program + platform.
    • Anything to do with Mox maybe?
  • We built the LTFF Grantee Dashboard
  • We worked with ACX Grants, will probably do that again.

  RESPONSE:

  • We built the LTFF Grantee Dashboard
  • As mentioned previously we worked with ACX on their Grants program and will probably do soon again.
  • We are generally interested in increasing our collaboration with other orgs, such as referring promising projects to them. This has happened informally so far, with funders such as OpenPhil and SFF making grants to projects that first gained visibility through Manifund.

• Period for requesting funding?

  • 1 year (?)

• Expected budget for that period (in USD)?

  • $400K/year maybe? If just Manifund ops, not including Mox, Manifest, etc.?

• Amount you are requesting from us (in USD)?

  • $200K

• What would those funds be spent on and why is that important?

  • Providing additional runway for the organization to cover current costs e.g. salary of our 1 FTE. We currently have enough funds for a little over a year.
  • We want to expand our AI safety regranting program and branch out into other projects, including evaluating giving opportunities in AI Safety and a trading fund for our DAF. This may necessitate bring on an additional FTE, e.g. an operations person or a writer.
  • [More later on why these projects are important specifically and why we need the funds for them]

RESPONSE: This money would primarily be spent on employment costs for our sole FTE (which we can currently continue for another ~18 months) and hiring a second FTE for operations and engineering. This is important for our ability to continue facilitating effective giving, as currently Austin Chen (co-founder & CEO of Manifund) is bottlenecked by time spent on general ops, website maintenance, etc., and this is preventing him and Manifund’s one FTE from quickly scaling up our existing programs, specifically AI safety regranting. More specifically, some of the work that these funds would enable includes:

• Fundraising for the regranting program (specifically trying to bring on more HNW individuals).
• Active management of the regranting program, e.g., onboarding regrantors, supporting collaboration and communication between regrantors, evaluating regrants, topping up regranting budgets.
• Improving Manifund’s service for donors, through the unique offering of a “prediction market trading fund” that will generate super-normal expected returns for money held in Manifund accounts.
• Publishing more detailed research and analysis of AI safety giving opportunities.

Additionally, part of what makes Manifund an effective organization is our flexibility and willingness to proactively take risks on new ideas and projects. This funding would give us some financial security to exercise that flexibility and pursue our comparative advantage — thus far we have needed to rely on the backstop of Austin’s personal finances.

• Provide an 80% confidence interval for the funds you expect to raise in that period (in USD), and explain in a few words why you think that estimate is reasonable.
  • God I really don’t know, maybe [$1.5M, $12M] per year?
  • Lower bound is roughly what we’ve been raising for the regranting program so far, and we intend to expand it this year. Upper bound is hard to judge, partially based on how much other orgs raise, partially based on the net worth of mine+Austin’s networks weighted by likelihood of giving.

RESPONSE:

Our 80% confidence interval for the effective giving we facilitate in the next year (April 2025 - April 2026) is [$2.5M, $30M]. We are pretty confident about the lower bound, as we have raised similar amounts in 2023 and 2024, and have so far raised $2.25M in 2025 for this year’s regranting program. Through our relationships with HNW donors, our affiliation with ACX Grants, and the rate of small-scale individual giving, we think it is likely but not guaranteed that we funnel >$2.5M to effective charities and projects in the next year-to-date. As for the upper bound, we think that there is a lot of potential upside due to a few factors:

• The significantly increased effort we plan to expend on fundraising, especially if we are able to become un-bottlenecked on operations & engineering.