Fortunately, GiveWell has agreed for me to post their response to my post on Independent Evaluation for Reputation; you can find it at the end of the post.
stevenhuynđ¸
Karma: 157
stevenÂhuynâs Quick takes
Reaping the benefits of AGI later is pretty insignificant in my opinion. If we get aligned AGI utopia, we will have utopia for millions of years. Acceleration by a few years if negligible if it increase p(doom) by >1%.
This is not true depending on what you think AGI utopia will look like. Thereâs some math outlined in What We Owe the Future about this dilemma i.e. area under the curve of these hypothetical AGI utility functions.
Getting utopia 1 year faster creates a 2x better universe. (hypothetically)
I hope so! Apparently the concept was received well with the team.
GiveWellâs Carley Moor from their philantrophic outreach team contacted me and we had a conversation a few weeks ago which prompted this post.
Among other things I asked about independent verification there. The short answer seems to be no independent verification with the caveat that they adjust. The spreadsheets I linked were sourced from her.
They do fund at least one meta charity that help improve monitoring & evaluation at these charities.
I asked her to either post her response email here or let me post it verbatim and am awaiting to hear from her next week. Being cautious lest I misrepresent them.
Transparency is only a means for reputation. The world is built on trust and faith in the systems and EA is no different.
I believe more people would be alarmed by the lack of independent vetting than the nominal cost effective numbers being inaccurate themself. It feels like there are perverse incentives at play.
Appreciate the feedback, although can you elaborate on what you mean by impact data and progamattic data?
I agree I could have made a better case on the reputation part.
This is what I meant.
RepÂuÂtaÂtion HarÂdenÂing: GiveWell
Reputation Hardening
Prompted largely by the fall in EA credibility in recent years. And also being unsatisfied with GiveWellâs lack of independent verification of the charities they recommend.
Here is a lightly edited AI generated slop version:
Reputation Hardening: Should GiveWell Verify Charity Data Independently?
âReputation hardeningâ involves creating more resilient reputations.
Recent events have shown how reputation damage to one EA entity can affect the entire movementâs credibility and therefore funding and influence. While GiveWellâs evaluation process is thorough, it largely relies on charity-provided data. I propose they consider implementing independent verification methods.
Applying to GiveWell/âGHD
Mystery evaluators conducting unannounced site visits
Independent surveyors (potentially hiring former charity employees) collecting impact data
Creating funding opportunities for third-party monitoring organizations GiveWell have already been involved in market shaping
These measures could help detect potential issues early and strengthen confidence in effectiveness estimates.
This is a preliminary idea to start discussion. What other verification methods or implementation challenges should we consider?
I see, appreciate the responses
Definitely an assumption of mine: but those with kidney issues would already have to be educated on what foods/âminerals to avoid I had imagined.
Would you happen to know any other subgroups?
btw Iâm a fan, your vid helped get me sign the pledge last year!
Supplements with a U-shaped benefit/âharm curve like that and different effects in different subgroups arenât appropriate for universal supplementation.
Is this a like a medical rule of thumb?
Iâd just imagine that all the other commonly fortified minerals and vitamins are u-shaped in outcomes. Calcium, iodine causing hyperthyroidism. Itâd just depend on the risks.
I did already glance at how likely potassium would harm others and you might find the current information interesting:
The studies suggest that so far it seems quite safe. However there could be a mild (1-2) point increase in blood pressure at the low doses for people who arenât hypertensive, which shouldnât impose much risk.
I misread:
Itâd depend on the food in question. The analysis however was under the assumption of it being additive, i.e. typical fortification.
Iâm not a baker or a cow farmer so I donât know in what situations salt can be replaced with potassium chloride. I think in a lot of cases potassium chloride should be able to be added with limited issues as saltier things tend to taste better.
NapÂkin Math AnalÂyÂsisâPoÂtasÂsium forÂtifiÂcaÂtion to reÂduce Blood PresÂsure
If you really felt bad, you would also have to be diligently doing research on suffering rates per calorie of each plant food.
Itâs not immediately obvious whether the crop deaths of a slice of bread which is more easily understood as vegan, causes less suffering than eating a farmed oyster and the killing of its resulting by-catch (barnacles perhaps?).
If anyone can point me towards any research of different foods â suffering (maybe neurons as a proxy?) I would love it.
An excerpt from your website. (which I love)
we aim to teach an entire generation to care a little bit more than the generations before them and to truly have an impact on the world, through the actions that we inspire. We are making kindness viral!
What are your insights on your main target audience generation alpha/âzoomer? What are your quantitative goals and/âor how do you measure your success?
I was suggesting in both.
Cool topic.
This is the key one to meditate on.
For me at least signing the giving pledge was a year of internalising that I have these values and I must eat them. Otherwise these arenât my values after all. Likewise for the standards of being a good friend, father, flutist etc.