All of TylerMaule's Comments + Replies

Thanks! Small correction: Animal Welfare YTD is labeled as $53M, when it looks like the underlying data point is $17M (source and 2023 full-year projections here)

Both posts contain a more detailed breakdown of inputs, but in short:

1. 80k seems to include every entry in the Open Phil grants database, whereas my sheet filters out items such as criminal justice reform that don't map to the type of funding I'm attempting to track.
2. They also add a couple of 'best guess' terms to estimate unknown/undocumented funding sources; I do not.

If you expect to take in $3-6M by the end of this year, borrowing say $300k against that already seems totally reasonable.

Not sure if this is possible, but I for one would be happy to donate to LTFF today in exchange for a 120% regrant to the Animal Welfare Fund in December^[1]

1. ^{^}

   This would seem to be an abuse of the Open Phil matching, but perhaps that chunk can be exempt

the comparison-in-practice I'm imagining is (say) $100k real dollars that we're aware of now vs $140k hypothetical dollars

That is very different from the question that Caleb was answering—I can totally understand your preference for real vs hypothetical dollars.

So these are all reasons that funding upfront is strictly better than in chunks, and I certainly agree. I'm just saying that as a donor, I would have a strong preference for funding 14 researchers in this suboptimal manner vs 10 of similar value paid upfront, and I'm surprised that LTFF doesn't agree.

Perhaps there are some cases where funding in chunks would be untenable, but that doesn't seem to be true for most on the list. Again, I'm not saying there is no cost to doing this, but if the space is really funding-constrained as you say 40% of value is an awful lot to give up. Is there not every chance that your next batch of applicants will be just as good, and money will again be tight?

A quick scan of the marginal grants list tells me that many (most?)^[1] of these take the form of a salary or stipend over the course of 6-12 months. I don't understand how the time-value of money could be so out of whack in this case—surely you could grant say half of the requested amount, then do another round in three months once the large donors come around?^[2]

1. ^{^}

   As for the rest, I don't see anything on the list that wouldn't exist in three months.

2. ^{^}

   Daniel's comment says "there are a whole host of issues" with this approach. I'd be curious to know wha

IDK 160% annualized sounds a bit implausible. Surely in that world someone would be acting differently (e.g. recurring donors would roll some budget forward or take out a loan)?

I would be curious to hear from someone on the recipient side who would genuinely prefer $10k in hand to $14k in three months' time.

Regarding the funding aspect:

• As far as I can tell, Open Phil has always given the majority of their budget to non-longtermist focus areas.
  • This is also true of the EA portfolio more broadly.
• GiveWell has made grants to less established orgs for several years, and that amount has increased dramatically of late.

IMHO seems possible to be rigorous with imaginary money, as some are with prediction markets or fantasy football. Particularly so if the exercise feels critical to the success of the platform.

I think the site looks great btw, just pushing back on this :)

Could you not dogfood just as easily with $50 (or fake money in a dev account)?

You may find this spreadsheet useful for that type of information

Substantially less money, through a combination of Meta stock falling, FTX collapsing, and general market/crypto downturns^[3]...[3] Although Meta stock is back up since I first wrote this; I would be appreciative if someone could do an update on EA funding

Looking at this table, I expect the non-FTX total is about the same^[1]—I'd wager that there is more funding commited now than during the first ~70% of the second wave period.^[2]

I think most people have yet to grasp the extent to which markets have bounced back:

• The S&P 500 Total Return Index is within

I second these suggestions. To get more specific re cause areas:

• Each source uses a different naming convention (and some sources are just blank)
• I'd suggest renaming that column 'labels' and instead mapping to just a few broadly defined buckets which add up to 100%—I've already done much of that mapping here

Borrowing money if short timelines seems reasonable but, as others have said, I'm not at all convinced that betting on long-term interest rates is the right move. In part for this reason, I don't think we should read financial markets as asserting much at all about AI timelines. A couple of more specific points:

Remember: if real interest rates are wrong, all financial assets are mispriced. If real interest rates “should” rise three percentage points or more, that is easily hundreds of billions of dollars worth of revaluations. It is unlikely that shar

I'm definitely not suggesting a 98% chance of zero, but I do expect the 98% rejected to fare much worse than the 2% accepted on average, yes. The data as well as your interpretation show steeply declining returns even within that top 2%.

I don't think I implied anything in particular about the qualification level of the average EA. I'm just noting that, given the skewedness of this data, there's an important difference between just clearing the YC bar and being representative of that central estimate.

A couple of nitpicky things, which I don't think change the bottom line, and have opposing sign in any case:

1. In most cases, quite a bit of work has gone in prior to starting the YC program (perhaps about a year on average?) This might reduce the yearly value by 10-20%
2. I think the 12% SP500 return cited is the arithmetic average of yearly returns. The geometric average, i.e. the realized rate of return should be more like 10.4%

I worry that this presents the case for entrepreneurship as much stronger than it is^[1]

1. The sample here is companies that went through Y-Combinator, which has a 2% acceptance rate^[2]
2. As stated in the post, roughly all of the value comes from the top 8% of these companies
3. To take it one step further, 25% of the total valuation comes from the top 0.1%, i.e. the top 5 companies (incl. Stripe & Instacart)

So at best, if a founder is accepted into YC, and talented enough to have the same odds of success as a random prior YC founder, $4M/yr might be a reasonable... (read more)

Yeah I think we're on the same page, my point is just that it only takes a single digit multiple to swamp that consideration, and my model is that charities aren't usually that close. For example, GiveWell thinks its top charities are ~8x GiveDirectly, so taken at face value a match that displaces 1:1 from GiveDirectly would be 88% as good as a 'pure counterfactual'

1. Most matches are of the free-for-all variety, meaning the funds will definitely go to some charity, just a question of who gets there first (e.g. Facebook & Every.org). While this might sound like a significant qualifier, it's almost as good as a pure counterfactual unless you believe that all nonprofits are ~equally effective.
   1. The 'worst case' is a matching pool restricted to one specific org, where presumably the funds will go there regardless, and doesn't really add anything to your donation.
2. Conversely, as Lizka noted, even the best counterfactual on

There are still funds remaining, but it looks like each person can only set up three matched donations

A lot of this wouldn't show up in malaria, e.g. last year 39% of GiveWell funds directed went to malaria programs. But yeah, still would be interested to see data.

Naively, $1.6B/$5k ~330k deaths averted^[1]? Adjust down because some spending is less effective than AMF. Adjust up because of AMF cost/life inflation.

1. ^{^}

   (Or equivalent)

It seems like it would be particularly difficult to know ahead of time whether one is well-suited to founding a charity, and I can imagine that is a major barrier to application. Do you have any suggestions for assessment of fit?

UK version

The biggest factor is the arrival of FTX, which has given more to infrastructure YTD than all others combined the prior two years

Relevant excerpt from his prior 80k interview:

Rob Wiblin: ...How have you ended up five or 10 times happier? It sounds like a large multiple.

Will MacAskill: One part of it is being still positive, but somewhat close to zero back then...There’s the classics, like learning to sleep well and meditate and get the right medication and exercise. There’s also been an awful lot of just understanding your own mind and having good responses. For me, the thing that often happens is I start to beat myself up for not being productive enough or not being smart enough or

Yes, sorry, on reflection that seems totally reasonable

Yeah it looked like grants had been announced roughly through June, so the methodology here was to divide by proportion dated Jan-Jun in prior years (0.49)

2012-Pres. (first longtermist grant was in 2015) no projection

Estimates for Open Phil:

FTX has so far granted 10x more to AI stuff than OPP

This is not true, sorry the Open Phil database labels are a bit misleading. 

It appears that there is a nested structure to a couple of the Focus Areas, where e.g. 'Potential Risks from Advanced AI' is a subset of 'Longtermism', and when downloading the database only one tag is included. So for example, this one grant alone from March '22 was over $13M, with both tags applied, and shows up in the .csv as only 'Longtermism'. Edit: this is now flagged more prominently in the spreadsheet.

Many of the sources used here can't be automated, but the spreadsheet is simple to update

Fixed

EA does seem a bit overrepresented (sort of acknowledged here).

Possible reasons: (a) sharing was encouraged post-survey, with some forewarning (b) EAs might be more likely than average to respond to 'Student Values Survey'?

I strongly agree with this comment, especially the last bit.

In line with the first two paragraphs, I think the primary constraint is plausibly founders [of orgs and mega-projects], rather than generically 'switching to direct work'.

Re footnote, the only public estimate I've seen is $400k-$4M here, so you're in the same ballpark.

Personally I think $3M/y is too high, though I too would like to see more opinions and discussion on this topic.

I enjoyed this post and the novel framing, but I'm confused as to why you seem to want to lock in your current set of values—why is current you morally superior to future you?

Do I want my values changed to be more aligned with what’s good for the world? This is a hard philosophical question, but my tentative answer is: not inherently – only to the extent that it lets me do better according to my current values.

Speaking for myself personally, my values have changed quite a bit in the past ten years (by choice). Ten-years-ago-me would likely be doing somethi... (read more)

Some of the comments here are suggesting that there is in fact tension between promoting donations and direct work. The implication seems to be that while donations are highly effective in absolute terms, we should intentionally downplay this fact for fear that too many people might 'settle' for earning to give.

Personally, I would much rather employ honest messaging and allow people to assess the tradeoffs for their individual situation. I also think it's important to bear in mind that downplaying cuts both ways—as Michael points out, the meme that direct ... (read more)

See also

Offsetting the carbon cost of going from an all-chicken diet to an all-beef diet would cost $22 per year, or about 5 cents per beef-based meal. Since you would be saving 60 chickens, this is three chickens saved per dollar, or one chicken per thirty cents. A factory farmed chicken lives about thirty days, usually in extreme suffering. So if you value preventing one day of suffering by one chicken at one cent, this is a good deal.

I didn't read the goal here as literally to score points with future people, though I agree that the post is phrased such that it is implied that future ethical views will be superior.

Rather, I think the aim is to construct a framework that can be applied consistently across time—avoiding the pitfalls of common-sense morality both past and future.

In other words, this could alternatively be framed as 'backtesting ethics' or something, but 'future-proofing' speaks to (a) concern about repeating past mistakes (b) personal regret in future.

I was especially interested in a point/thread you mentioned about people perceiving many charities as having similar effectiveness and that this may be an impediment to people getting interested in effective altruism

See here

A recent survey of Oxford students found that they believed the most effective global health charity was only ~1.5x better than the average — in line with what the average American thinks — while EAs and global health experts estimated the ratio is ~100x. This suggests that even among Oxford students, where a lot of outreach has b

Thanks for this—I have often wished I had a better elevator pitch for EA.

One thing I might add is some mention of just how wide the disparity can be amongst possible interventions, since this seems to be one of the most overlooked key ideas.

I believe both this post and Ben’s original ‘Funding Overhang’ post mentioned that this is an update towards a career with direct impact vs earning-to-give.

But earning-to-give is still very high impact in absolute terms.

Thanks for writing; I too have worried that many folks got the wrong impression here.

Anyone know where the $250k is coming from? This is all I could find:

Matching funds are provided by generous donors who contribute to help amplify grassroots giving

Is there any consideration for Investing-to-Give in the survey?

1. Is contributing to a DAF meant to count as a 'donation'? I would think yes, though not all I2G is done via DAF
2. According to this post, many in the community think deploying <5% of available capital/yr is currently optimal

Perhaps it could be interesting to ask for both 'amount donated' and 'amount earmarked for donation'?

Depends immensely on if you think there are EAs who could start billion-dollar companies, but would not be able to without EA funding. I.e. they're great founders, but can't raise money from VCs.

I think the core argument here is that not enough EAs try to start a company, as opposed to try and are rejected by VCs. IMO the point of seeding would be to take more swings.

Also, presumably the bar should be lower for an EA VC, because much of the founders' stake will also go to effective charity.