The impact purchase has so far distributed $5k, out of $10k planned for 2015.

(If you don't know what the impact purchase is: the original announcement is here; my post on "certificates of impact" is here; Ben Kuhn posted a better explanation here; past results are here.) 

We’re going to spend the final $5k in December. The last round is going to be significantly simplified, both for applicants and for us. Here’s how it will work:

  1. Sellers may submit a one-sentence description of what they did, an optional link, and a price at which they would be willing to "sell" some portion of the impact. This process should take only a few minutes.
  2. We may ask further questions or further investigate the particular projects that we are most likely to purchase.
  3. We, or other buyers, will accept some of the offers made in step [1]. If a price is attractive but not quite good enough then we may make a counteroffer, and haggling may ensue. The offers in step [1] aren't formally binding, so the seller can still back out (though we'd appreciate if you don't, since doing so wastes time).
If you want to submit an offer, you can either make a comment on this post, or fill out the form here. If you've submitted a project to a previous round there is no need to resubmit (though we are likely to email you to elicit your new asking price, given that the structure of the auction has changed).

The key change is that we are no longer trying to incentivize honesty. We realized that trying to be incentive-compatible made the process much more expensive, for both us and the sellers.  We also realized that the incentives were never going to be perfect or even great, once we had a dynamic process with more than one buyer.

My guess is that blog posts (or other relatively self-contained writings) are the most promising candidates; I recommend a low bar for quickly submitting a link and a high-ball estimate for the value of a post, as long as you won't be offended when we don't fund it. 

As always, our evaluations will be unapologetically arbitrary, rough, and biased.

The deadline for submissions is Sunday December 20.

4

0
0

Reactions

0
0
Comments9


Sorted by Click to highlight new comments since:

The idea that a fund outlay needs to be spent in haste since its completion date is approaching reminds me of the perverse incentives created in academic grants. What are your thoughts on simply rolling over the unspent money to increase the money you'll grant in 2016, rather than spending it in a hurry?

Several of the most promising applicants from previous rounds were willing to roll their applications to the new format. It's not clear that haste makes significant (any?) waste. We also weren't getting many new applicants per month without any new activity on our part, and we are generally interested in conserving time.

Since we never announced how much we would spend in 2016, it's hard to demonstrate that extra money is really extra.

Why are you paying full price, instead of trying to buy them at a discount? Or has that changed? Like let's suppose that each project you fund becomes 10% more likely because of the certificate scheme. In this situation, you should be trying to buy a certificate for the entire project at 10% of the altruistic value.

I disagree with your use of "should."

We also realized that the incentives were never going to be perfect or even great, once we had a dynamic process with more than one buyer.

Surely having multiple buyers improves incentive compatibility? In the limit consider the stock market, with a arbitrarily large number of buyers and very well aligned incentives.

Competition amongst buyers improves incentives in general, but the added complexity quickly destroys the specific guarantee of incentive-compatibility enjoyed by a second price auction.

For example, when the sale price in one round is published, it means that making a cheap sale can easily be worse than making no sale at all, since it affects what you might be able to earn in future rounds.

That's a particularly hard example to avoid. There are a bunch of other issues that seem easier to avoid, but only by placing increasingly strict constraints on the behavior of every buyer (e.g. once you make an offer on an item, you can never make any lower offer on the same item).

What's the deadline for submitting?

December 20, sorry for leaving that out of the original post.

It's excellent to see you complete this! Good luck. Do you plan to write-up an evaluation of the project, and your thoughts on the best ways to ensure good EA work gets funded?

Curated and popular this week
 ·  · 11m read
 · 
Confidence: Medium, underlying data is patchy and relies on a good amount of guesswork, data work involved a fair amount of vibecoding.  Intro:  Tom Davidson has an excellent post explaining the compute bottleneck objection to the software-only intelligence explosion.[1] The rough idea is that AI research requires two inputs: cognitive labor and research compute. If these two inputs are gross complements, then even if there is recursive self-improvement in the amount of cognitive labor directed towards AI research, this process will fizzle as you get bottlenecked by the amount of research compute.  The compute bottleneck objection to the software-only intelligence explosion crucially relies on compute and cognitive labor being gross complements; however, this fact is not at all obvious. You might think compute and cognitive labor are gross substitutes because more labor can substitute for a higher quantity of experiments via more careful experimental design or selection of experiments. Or you might indeed think they are gross complements because eventually, ideas need to be tested out in compute-intensive, experimental verification.  Ideally, we could use empirical evidence to get some clarity on whether compute and cognitive labor are gross complements; however, the existing empirical evidence is weak. The main empirical estimate that is discussed in Tom's article is Oberfield and Raval (2014), which estimates the elasticity of substitution (the standard measure of whether goods are complements or substitutes) between capital and labor in manufacturing plants. It is not clear how well we can extrapolate from manufacturing to AI research.  In this article, we will try to remedy this by estimating the elasticity of substitution between research compute and cognitive labor in frontier AI firms.  Model  Baseline CES in Compute To understand how we estimate the elasticity of substitution, it will be useful to set up a theoretical model of researching better alg
 ·  · 6m read
 · 
TLDR: This 6 million dollar Technical Support Unit grant doesn’t seem to fit GiveWell’s ethos and mission, and I don’t think the grant has high expected value. Disclaimer: Despite my concerns I still think this grant is likely better than 80% of Global Health grants out there. GiveWell are my favourite donor, and given how much thought, research, and passion goes into every grant they give, I’m quite likely to be wrong here!   What makes GiveWell Special? I love to tell people what makes GiveWell special. I giddily share how they rigorously select the most cost-effective charities with the best evidence-base. GiveWell charities almost certainly save lives at low cost – you can bank on it. There’s almost no other org in the world where you can be pretty sure every few thousand dollars donated be savin’ dem lives. So GiveWell Gives you certainty – at least as much as possible. However this grant supports a high-risk intervention with a poor evidence base. There are decent arguments for moonshot grants which try and shift the needle high up in a health system, but this “meta-level”, “weak evidence”, “hits-based” approach feels more Open-Phil than GiveWell[1]. If a friend asks me to justify the last 10 grants GiveWell made based on their mission and process, I’ll grin and gladly explain. I couldn’t explain this one. Although I prefer GiveWell’s “nearly sure” approach[2], it could be healthy to have two organisations with different roles in the EA global Health ecosystem. GiveWell backing sure things, and OpenPhil making bets.   GiveWell vs. OpenPhil Funding Approach What is the grant? The grant is a joint venture with OpenPhil[3] which gives 6 million dollars to two generalist “BINGOs”[4] (CHAI and PATH), to provide technical support to low-income African countries. This might help them shift their health budgets from less effective causes to more effective causes, and find efficient ways to cut costs without losing impact in these leaner times. Teams of 3-5
 ·  · 3m read
 · 
We’re excited to announce SparkWell! What is SparkWell? SparkWell is an Anti Entropy program designed to help high-impact nonprofit projects test ideas, develop operational capabilities, and launch as independent entities. We provide a temporary home for a diverse range of promising initiatives. Why have we built this? We believe that we’re living through a transformational period in history. Catastrophic risks loom large, whether from climate change, factory farming, pandemics, nuclear or cyber warfare, or the misalignment or misuse of intelligent systems. A transformational period in history warrants a transformation in philanthropy — and we want to give innovative projects the support they need to test their ideas and scale. We leverage our skills and experience with nonprofit operations to guide enrolled projects through a bespoke acceleration roadmap.  Within 6–24 months, your project will graduate into an independent entity with operational best practices. This will put you in a position to scale your activities — and help mitigate the catastrophic risks facing us. What does SparkWell offer? SparkWell offers 6-month, 12-month, or 24-month tracks to accommodate projects at different stages. We enable each project to: * Test ideas * Receive tax-exempt funding via Anti Entropy's 501(c)(3) * Run your project, including hiring staff, contractors, and managing expenses * Receive feedback and develop your theory of change * Develop operational capabilities * Access your bank account, company card, and dashboard * Receive mentorship and resources from your Project Liaison * Leave bookkeeping and compliance to us * Launch an independent entity * Monitor your progress along entity formation milestones * Be on track to independence within 6, 12, or 24 months * Launch an independent entity when you’re ready We apply a 7% service fee on funds raised or received during the program. You can learn more about the program here. Who are