Hide table of contents

Inducement prizes are cash prizes awarded to people who accomplish a particular feat specified ahead of time. Inducement prizes offer advantages over traditional hiring practices since the prize is allocated according to a post-hoc evaluation of performance, rather than upfront to a specific worker. Here, I propose a specific model for a small inducement platform intended to facilitate the creation of high quality effective altruism research.

The problem

Inducement prizes have experienced a long history; accordingly, economics literature going back over a century has provided empirical evidence that inducement prizes can spur innovation more efficiently than hiring researchers or engineers directly in certain circumstances (for some examples see this paper and this one).

Many existing organizations, such as the X Prize Foundation, have enjoyed some success by facilitating the creation of large inducement prizes. However, most inducement prizes are not facilitated on content-specific platforms. Rather, companies, non-profits, governments, and rich individuals interested in creating inducement prizes generally go through their own media, such as for the Brain Preservation Technology Prize and the Methuselah Mouse Prize.

Facilitating your own inducement prize contest makes sense for large bounties, but I'm currently unaware of any specific platforms dedicated to the procurement of smaller prizes, such as those with sums less than $10,000.

The closest platform that I'm currently aware of is the private Facebook group Bountied Rationality. However, there are a number of problems I see with Bountied Rationality, which should provide some indication for how I think an alternative platform could be improved,

  • The group visibility is set to private, which makes it harder to share research value created inside the group.
  • The integrity of the group itself is held up by the personal trust of those within the rationalist/effective altruism community, rather than a market-driven model of reputation.
  • There is no means of arbitrating disputes, and therefore there is no guarantee that people will be paid fairly for accomplishing the task as specified.

I still think that the group Bountied Rationality creates a lot of value. But the issues I've outlined above plausibly limit its ability to grow larger, and hamper its status as a reliable engine for producing outsourced insights and research.

The proposal

My proposed alternative is a public, market-driven bounty system aimed at procuring small inducement prizes, targeted at the effective altruism community. Below, I'll list a specific set of features which I think could help the platform to thrive.

Public

The first main difference between my model and the group Bountied Rationality is that content on the platform would be public. This feature makes the platform less suitable for personal requests, but more suitable for public research, such as inducing mathematics results, well-crafted bibliographies, well-sourced research summaries for a given topic, and in-depth investigations into potential interventions.

The Effective Altruist Forum, Lesswrong and Stackexchange already allow for something similar, in that they allow users to ask public questions, and the community answers are then curated via upvotes and downvotes. This model has been helpful to many, but in my experience, people are often hesitant to provide long-form and well-sourced answers to questions, probably due to the lack of strong incentives offered to those who give good answers.

Escrow and arbitration

The second main feature I propose is a requirement that people put their money in escrow, and that they must name someone as the arbitrator for the bounty. Escrow ensures that the bounty offerers cannot simply keep their money long after someone has already satisfied the conditions of the bounty (a problem with which I have personally been acquainted with).

The purpose of requiring people to name someone as an arbitrator serves a similar purpose as escrow. By naming a trusted third party to settle disputes, bounty offerors would be encouraged outline very specific conditions under which they want their bounty to be distributed. This incentive, and the fact that the offerors cannot simply unfairly refuse to pay, provides bounty hunters assurance that they will get paid if they perform the task successfully.

My own experience on Metaculus made me realize just how important it is for platforms to build solid mechanisms to build community trust. Even though people on Metaculus are not trading with real money, disputes can become agonizing and people can get angry when questions do not resolve in the way that they thought it would. As a result of these issues, Metaculus moderators and admins have become very careful in the way that they write questions, to ensure that questions are resolved unambiguously whenever possible.

One market-driven way of ensuring community trust is to openly allow users to bid to become arbitrators of particular bounties. In effect, the role of arbitrator could be something like a paid position: they would provide the services of trust and reliability, which could then flow through the platform, promising a fair environment for the bounty offerors and bounty hunters.

Targeted at the effective altruism community

My limited research suggests that some bounty-like services already exist. The most common bounty platforms are bug-bounty platforms, which at the moment far esclipse the size of the Facebook group Bountied Rationality.

However, even as some of these platforms exist -- and perhaps even one exists that uses the arbitration system I described above -- a large potential drawback comes from networking effects. If an EA tried to induce complex research using an existing platform, they would be unlikely to attract the people best suited to doing that research. As a result, an EA would be better off just trying to induce the research more informally, either by asking for people to collaborate with them in the community, or by hiring someone to perform the research directly.

My hope is that creating a platform that facilitates small-scale inducement prize contests would help solve this problem, better allowing EAs in need of research solutions to target people most likely to provide them.

Comments10


Sorted by Click to highlight new comments since:

I am pretty excited about the potential for this idea, but I am a bit concerned about the incentives it would create. For example, I'm not sure how much I would trust a bibliography, summary, or investigation produced via bounty. I would be worried about omissions that would conflict with the conclusions of the work, since it would be quite hard for even a paid arbitrator to check for such omissions without putting in a large amount of work. I think the reason this is not currently much of a concern is precisely because there is no external incentive to produce such works - as a result, you can pretty much assume that research on the Forum is done in good faith and is complete to the best of the author's ability.

Potential ways around this that come to mind:

  • Maybe linking user profiles on this platform to the EA Forum (kind of like the Alignment Forum and LessWrong sharing accounts) would provide sufficient trust in good intentions?
  • Maybe even without that, there's still such a strong self-selection effect anyway that we can still mostly rely on trust in good intentions?
  • Maybe this only slightly limits the scope of what the platform can be used for, and preserves most of its usefulness?

Potential ways around this that come to mind:

Good ideas. I have a few more,

  • Have a feature that allows people to charge fees to people who submit work. This would potentially compensate the arbitrator who would have to review the work, and would discourage people from submitting bad work in the hopes that they can fool people into awarding them the bounty.
  • Instead of awarding the bounty to whoever gives a summary/investigation, award the bounty to the person who provides the best  summary/investigation, at the end of some time period. That way, if someone thinks that the current submissions are omitting important information, or are badly written, then they can take the prize for themselves by submitting a better one.
  • Similar to your first suggestion: have a feature that restricts people from submitting answers unless they pass certain basic criteria. E.g. "You aren't eligible unless you are verified to have at least 50 karma on the Effective Altruist Forum or Lesswrong." This would ensure that only people from within the community can contribute to certain questions.
  • Use adversarial meta-bounties: at the end of a contest, offer a bounty to anyone who can convince the judge/arbitrator to change their mind on the decision they have made.

another incentive system/component I have seen is that forums will allow users not only to upvote but to give other incentives to good answers. stackoverflow has bounty, and reddit coins

I would find this useful. But:

- What is the likely market size for this platform?
- How much would it cost to develop? (Including escrow infrastructure?)
- What would the fees for use be / need to be to keep the platform afloat?

What is the likely market size for this platform?

I'm not sure, but I just opened a Metaculus question about this, and we should begin getting forecasts within a few days. 

Briefly:

  1. I like the idea
  2. Think it will work
  3. Also like the idea of using Metaculus to forecast this

Somewhat related, and potentially relevant if someone sets this up:

  1. The Nonlinear Fund wrote up why they use RFPs (Requests For Proposals). 
  2. Certificates of Impact.
  3. There is an upcoming project platform for EAs, designed to coordinate projects with volunteers. A forum post should be out soon, but meanwhile you can see a prototype here.

Thank you for this great post. In the past I have looked for such platforms and concepts, but was unaware of the term 'inducement prize' and did not find much. 

Two extensions to the concept you presented could make it even more interesting, especially for the EA community. Firstly, rather than just requests being supplied to such a platform, offers to conduct e.g.  research  could be posted first by  qualified researchers in order to gauge interest. Secondly, there is no reason why there couldn't be several parties/individuals who pay the bounty collectively.  Essentially, this would be a "reverse kickstarter" use case, where  payment is made  after completion rather than in the beginning.  

It seems that there a lot of potential projects in the community with distributed interest and willingness-to-pay : literature reviews,  evaluations of possible cause areas,  research into personal Covid-19 risks etc. 

I think I'm not super optimistic about this idea, mainly because it seems like it's somewhat common for coordination platforms to be built but then people don't actually coordinate to all start using them. But: 

  • I've got a lot of uncertainty, and see "this is a great idea" as plausible
  • The cost of making an MVP seems probably low
  • So I think if someone's enthusiastic about making an MVP of this, it may well be worth doing so

(This is related to the general point that if you try something and it fails, you can stop putting resources into it, but if it works you can continue putting resources in and getting value out for a while.)

I've drafted a post (which I'll publish in ~3 weeks) proposing "A central, editable database to help people choose and do research projects". This would have something similar as one of its components. But it's possible that what I propose is overly complicated and it would be better to do one or more components from it in a separate, simplified way (which could then look like your proposal). In any case, anyone interested in a longwinded draft without an MVP can check it out (and leave comments if they want) here :)

I‘d also be really interested to see more attempts in this direction. I suspect that there are a many smaller research projects and people interested in working on such projects and this could being those together and result in interesting insights and learning opportunities.

Just in case you weren’t aware, LessWrong has tags on open and closed bounties that also might provide some interesting data: https://www.lesswrong.com/tag/bounties-closed https://www.lesswrong.com/tag/bounties-active

Curated and popular this week
 ·  · 12m read
 · 
Economic growth is a unique field, because it is relevant to both the global development side of EA and the AI side of EA. Global development policy can be informed by models that offer helpful diagnostics into the drivers of growth, while growth models can also inform us about how AI progress will affect society. My friend asked me to create a growth theory reading list for an average EA who is interested in applying growth theory to EA concerns. This is my list. (It's shorter and more balanced between AI/GHD than this list) I hope it helps anyone who wants to dig into growth questions themselves. These papers require a fair amount of mathematical maturity. If you don't feel confident about your math, I encourage you to start with Jones 2016 to get a really strong grounding in the facts of growth, with some explanations in words for how growth economists think about fitting them into theories. Basics of growth These two papers cover the foundations of growth theory. They aren't strictly essential for understanding the other papers, but they're helpful and likely where you should start if you have no background in growth. Jones 2016 Sociologically, growth theory is all about finding facts that beg to be explained. For half a century, growth theory was almost singularly oriented around explaining the "Kaldor facts" of growth. These facts organize what theories are entertained, even though they cannot actually validate a theory – after all, a totally incorrect theory could arrive at the right answer by chance. In this way, growth theorists are engaged in detective work; they try to piece together the stories that make sense given the facts, making leaps when they have to. This places the facts of growth squarely in the center of theorizing, and Jones 2016 is the most comprehensive treatment of those facts, with accessible descriptions of how growth models try to represent those facts. You will notice that I recommend more than a few papers by Chad Jones in this
LintzA
 ·  · 15m read
 · 
Introduction Several developments over the past few months should cause you to re-evaluate what you are doing. These include: 1. Updates toward short timelines 2. The Trump presidency 3. The o1 (inference-time compute scaling) paradigm 4. Deepseek 5. Stargate/AI datacenter spending 6. Increased internal deployment 7. Absence of AI x-risk/safety considerations in mainstream AI discourse Taken together, these are enough to render many existing AI governance strategies obsolete (and probably some technical safety strategies too). There's a good chance we're entering crunch time and that should absolutely affect your theory of change and what you plan to work on. In this piece I try to give a quick summary of these developments and think through the broader implications these have for AI safety. At the end of the piece I give some quick initial thoughts on how these developments affect what safety-concerned folks should be prioritizing. These are early days and I expect many of my takes will shift, look forward to discussing in the comments!  Implications of recent developments Updates toward short timelines There’s general agreement that timelines are likely to be far shorter than most expected. Both Sam Altman and Dario Amodei have recently said they expect AGI within the next 3 years. Anecdotally, nearly everyone I know or have heard of who was expecting longer timelines has updated significantly toward short timelines (<5 years). E.g. Ajeya’s median estimate is that 99% of fully-remote jobs will be automatable in roughly 6-8 years, 5+ years earlier than her 2023 estimate. On a quick look, prediction markets seem to have shifted to short timelines (e.g. Metaculus[1] & Manifold appear to have roughly 2030 median timelines to AGI, though haven’t moved dramatically in recent months). We’ve consistently seen performance on benchmarks far exceed what most predicted. Most recently, Epoch was surprised to see OpenAI’s o3 model achieve 25% on its Frontier Math
Omnizoid
 ·  · 5m read
 · 
Edit 1/29: Funding is back, baby!  Crossposted from my blog.   (This could end up being the most important thing I’ve ever written. Please like and restack it—if you have a big blog, please write about it). A mother holds her sick baby to her chest. She knows he doesn’t have long to live. She hears him coughing—those body-wracking coughs—that expel mucus and phlegm, leaving him desperately gasping for air. He is just a few months old. And yet that’s how old he will be when he dies. The aforementioned scene is likely to become increasingly common in the coming years. Fortunately, there is still hope. Trump recently signed an executive order shutting off almost all foreign aid. Most terrifyingly, this included shutting off the PEPFAR program—the single most successful foreign aid program in my lifetime. PEPFAR provides treatment and prevention of HIV and AIDS—it has saved about 25 million people since its implementation in 2001, despite only taking less than 0.1% of the federal budget. Every single day that it is operative, PEPFAR supports: > * More than 222,000 people on treatment in the program collecting ARVs to stay healthy; > * More than 224,000 HIV tests, newly diagnosing 4,374 people with HIV – 10% of whom are pregnant women attending antenatal clinic visits; > * Services for 17,695 orphans and vulnerable children impacted by HIV; > * 7,163 cervical cancer screenings, newly diagnosing 363 women with cervical cancer or pre-cancerous lesions, and treating 324 women with positive cervical cancer results; > * Care and support for 3,618 women experiencing gender-based violence, including 779 women who experienced sexual violence. The most important thing PEPFAR does is provide life-saving anti-retroviral treatments to millions of victims of HIV. More than 20 million people living with HIV globally depend on daily anti-retrovirals, including over half a million children. These children, facing a deadly illness in desperately poor countries, are now going