Hey Vasco! I just read the summary here, so my apologies if I’m missing something important.
I think I disagree with the line of reasoning you’re following. For instance, you say:
>>I estimated its HSI has been 0.0292 % as cost-effective as HIPF accounting for effects on the target beneficiaries, and soil ants, termites, springtails, mites, and nematodes.
It would be very surprising to me if we can know with any degree of confidence that exploratory research into highly speculative areas is >3000X more cost effective than Shrimp Welfare Project’s main program—a program which seems to be one of the most cost effective on the planet!
I don’t doubt that every step in specific chain of reasoning is mostly correct. Rather, I think where I get off the boat here is that I think the uncertainty in these flowthrough effects tends to dominate after a certain point. I get the sense you are willing to take these much more seriously and literally.
Another area of disagreement is the general framing of suggesting that we should always just fund the one most impactful thing. I agree that this makes sense on the margin, but I expect we’re generally more likely to get to a better world in the end if we take a portfolio approach and fund and encourage lots of projects that pass a certain bar. A large part of this is driven by the fact that you get seriously diminishing returns on projects (e.g. I’d be surprised if CEARCH could cost effectively deploy SWP’s full annual budget).
So I’d probably frame all of this as more “Both CEARCH’s program and SWP’s program perform very well on these metrics, and we should consider funding them.”
From Fish Welfare Initiative (FWI):
Updates:
We have launched new studies on feed fortification.
We posted updates on our various other studies, including on our use of satellite imagery to remotely monitor water quality.
We published our 2026 priorities and our 2025 year in review.
Open Roles:
We are hiring for an Operations Associate! Applicants from India and the Philippines are preferred, but international applicants are welcome. This is a good role for an early-career EA person. Deadline: Feb 20.