Gage Weston

273 karmaJoined Jun 2020Seeking workAustin, TX, USA


  • Attended an EA Global conference
  • Received career coaching from 80,000 Hours
  • Attended more than three meetings with a local EA group


Very good points made! One objection I think you didn’t mention that might be on OP’s mind in neartermist allocations has to do with population ethics. One reason many people are near termist is because they subscribe to a person-affecting view whereby the welfare of “merely potential” beings does not matter. Since basically all animal welfare interventions either 1. Cause fewer animals to exist, or 2. Change welfare conditions for entire populations of animals, it seems extremely unlikely the animals who would otherwise have lived the higher suffering lives will have the same identity (eg same genes) as the higher welfare ones. To a person affecting view, this implies animal welfare interventions like corporate campaigns or alt protein investment merely change who or how many animals there are but don’t benefit any animal in particular and thus have no impact on this moral view. I personally don’t subscribe to this view, and I am not sure if most people at OP with a person affecting view have taken this idea seriously although it does seem like the right conclusion from this view.

re conflict of interest concerns: I'd go ahead and apply for both membership as a donor and for funding for your project. Others may do the same. We are still considering how to weigh COI concerns against potential value of allowing those applications in, and plan to arrive at a more conclusive policy before deciding whether or not to admit members/applications. 

Re referring other applications: similar to above, though I think there's less conflict here and would perhaps unfairly limit many good applications if we excluded merely because you are well acquainted with an applicant, so I would encourage them to apply and encourage you to encourage them too :)

Re $100K minimum: you do not need to donate $100K+ yourself, as long as you are in one way or another influencing $100K+ that could reasonably be expected to go toward applications from our circle. So advisors or grant-makers moving money to meta things are okay, although we prefer to maintain more distance from OP and EA Funds grant-makers. It is also ok if you expect to move slightly below $100K/year or you are uncertain how much it'll be – it'll be more of a judgement call from our circle there as to who we admit.

Since we're just getting started, we expect much of our plans and policies will change, though.

We are connected to Nonlinear Network, but there is currently no interaction with them at least for this first round. We may interact with them in the future after our first round and depending on how well-aligned our goals end up being.

I'm curious why you and many EA's who focus on longtermism don't suggest donating to longtermist cause areas (as examples often focuses on Givewell or ACE charities). It seems like if orgs I respect like Open Phil and long term future fund are giving to longtermist areas, then they think that's among the most important things to fund, which confuses me when I then hear longtermists acting like funding is useless on the margin or that we might as well give to GiveWell charities. It gives me a sense that perhaps there's either some contradiction going on, or I'm missing something, but either way it makes it very difficult for me to get others excited about longtermism if they won't enter it with their career and even the die-hard longtermists are saying marginal funding is useless or at least worse than GiveWell charities.

While I agree there is a good signaling benefit, I think you need to be a bit more rigorous in figuring out just how impactful it is, and what the ultimate goal of signaling is. Taking your $100k/year example with Givewell's ~$5k/life saved, that'd mean that the signaling of one's donations below this amount are better than saving about 20 lives. That doesn't seem right to me... And if you think that signaling is valuable for community building, it's probably way more effective to just donate to community building (e.g. EA infrastructure fund) than anything else under your assumptions. I do donate to EAIF so I think community building is the most effective thing right now, but I don't think the main benefit of my donations is the signaling, I think it's the actual resources being moved. 

Most people won't get your signal since most donations just go into an org's bank account without some news headline or anything. There's probably much cheaper ways of signaling EA, like maybe wear a $15 EA t-shirt every day and save your $100k. (I'm not suggesting that's the most effective thing, but I'm just taking the signaling argument to its extreme).

I am now starting a book giveaway based on these data/arguments at my EA Austin fellowship, and I managed to buy a bunch of used versions of EA books for an average of $6-7 per book including shipping, so with a bit of thrift you can get some great deals! Amazon and eBay at least in US make buying used books quite cheap 

I have an idea to increase EA donation matches on Facebook's Giving Tuesday, and I want your feedback! PLEASE FILL OUT THIS GOOGLE FORM here and leave a comment after reading this.

I am considering creating and distributing an automated tool that can schedule and execute donations immediately when matching begins, even if the donor is asleep or AFK. Last year, about 50% of the $1.2m+ donated by EA's was matched. With this tool, this percentage could easily approach 100% (increasing match funds to EA causes by $500k+) since the main reasons people didn't get matched were 1. people couldn't click buttons within the 13 seconds it took for matches to be met, and 2. people didn't want to get up as early as 5 AM in parts of USA. People would likely be more willing to donate since the odds of getting matched are higher, and they don't have to incur the cost of waking up early or of practicing for the match.

However, I (and others I have spoken to) are worried this tool might have some negative outcomes that could outweigh the benefits, such as:

1. it could garner negative press for EA or the orgs that the tool's users donated to by displaying unfair practices since no one else is using automation.

2. It might violate facebook's terms, stating "You may not access or collect data from our Products using automated means", although I wouldn't consider this "accessing" or "collecting" data. There's also no statement FB's GT policies mentioning automation.

3. might not get matched by FB if the automation were detected

4. Non-EA's might find or recreate the tool and now we're no better off

5. Facebook might change GT rules or shut down GT if automation enters the game

I am considering the following options (for you to vote on in the form):

1. SCALE IT: Distribute the automation to EA community to use on 2020's Giving Tuesday at full scale
2. TEST IT: Only distribute it to a select number of people in 2020 in order to see if the tool works, if we get any backlash from FB or from the public, etc. Pending test results, we may then decide whether or not to SCALE IT in 2021
3. KILL IT: Do not distribute an automation at all.

In addition to voting on this, please let me know in the form what your thoughts are regarding the arguments for/against, if you have any new arguments/information that'd be relevant, and if you have any advice for how to best create and distribute the tool or message me if you want to help in this process (assuming we decide not to KILL IT!). The form answers will be open to all viewers in the form.