The Cheapest Mercy
In 2024 we slaughtered 87,896,729,120 land animals for food.1 About 76 billion of them were chickens. The rest were ducks, pigs, geese, sheep, goats, turkeys and cattle, in that order. We kill around 240 million land animals a day.2
That is the small number.
The large one is harder to write down, and it took me a while to find it. We count land animals to the last animal, but we count fish by weight. The FAO publishes the global catch in tonnes, not in fish.3 When we turn the tonnes back into animals by dividing the total weight by the weight of an average fish, the numbers get very large. Somewhere between 1.1 and 2.2 trillion wild fish are pulled from the oceans each year. Another 124 billion or so are farmed and killed in pens. Between 255 and 605 billion farmed crustaceans go with them, mostly shrimp.4 Add the farmed insects, on the order of a trillion a year, and the totals stop meaning much.5 Counted one creature at a time, the number of vertebrates and crustaceans we kill for food is somewhere around 4 to 8 billion a day.
Large numbers tend to switch the moral part of the brain off rather than on, so set the total aside and take one practice at its real size. Because male chicks cannot lay eggs, and the laying breeds grow too slowly to be worth raising for meat, the egg industry kills its males on the day they hatch, either by dropping them alive into a grinder or by gassing them. This happens to roughly 6.5 billion chicks a year.6 It is how egg production has been designed to work for decades, an ordinary line item, costed and scheduled like any other.
The usual reason for not minding is that these are only animals, and the distance between them and us is wide enough that what happens across it doesn’t really matter. That is getting harder to believe. The view that fish feel pain is now the scientific mainstream, and the evidence that some invertebrates feel it too, including the crabs and lobsters we boil alive, is strong enough that several governments have written them into their animal-welfare laws.7 Their nervous systems are built from the same parts as ours, including the parts that produce pain and fear. They are closer to us than we like to think.
I think a later century will look back at this the way we now look at things earlier centuries accepted as normal, and will struggle to understand how people who were decent in every other way sat inside a catastrophe this large and mostly thought about lunch. Every widening of the moral circle has looked obvious afterward and impossible at the time, and every one of them was slowed for generations by people who were sure the current edge was where it naturally stopped. In raw numbers this case may be larger than all of them.
I should be straight about my own position. I have been vegan / vegetarian for ~22 years now, and I have made the moral argument when I was younger more times than I can count. It does not work. Not at the scale that matters, anyway. Global meat production reached about 365 million tonnes in 2024 and is forecast to keep climbing for at least another decade, led by exactly the two categories, poultry and farmed fish, that kill the most individuals per tonne.8 The number of people who will read something like this and change what they eat is a rounding error against the demand growth of a single year in Asia. The moral case against all this has been available for 50 years. You can win that argument every time and still watch the numbers go up, which is what has happened.
So I am not going to ask anyone to go vegetarian, or try to reduce their overall consumption. Partly because I don’t think the asking does much, and partly because I have come to think it was never where the leverage was.
The leverage is in how moral progress actually happens. We like to believe the great moral advances were won by persuasion, by someone making the case so well that the public’s conscience finally turned. Persuasion matters, and the people who make the case early may matter enormously. But the majority does not move on persuasion. It moves when the better choice becomes the cheaper one. The killing in the numbers above does not continue because billions of people have weighed a chicken’s life against their dinner and decided the dinner wins. It continues because the alternative has cost more and tasted worse to many. Close that gap and the moral question answers itself, with no one having to win it, for the large majority who were never going to be argued out of a habit but will switch the moment switching is free.
Which is why the thing I find genuinely hopeful is not a change in values but a set of cost curves, and why the people that will end the largest cruelty in history are not the activists but the engineers and the companies that will never once think about the animal.
There are two families of technology that matter, and they sit at very different stages. The first is precision fermentation, and it is already real, and starting to be commercially available. You take a microbe (a yeast or a fungus), and give it the gene for a milk protein or an egg protein, and it secretes the exact molecule a cow or a hen would have made, with no cow or hen anywhere in the process. We have made insulin and cheese rennet this way for decades, and the industrial base for it is mature. Perfect Day and Remilk already sell fermentation-made dairy proteins in the United States and Singapore, in ice cream and cream cheese that contain the real thing without the animal.9 That is the proof the mechanism works: you can make the molecule people want and separate it from the body that used to be needed to grow it.
The second family is cultivated meat, real animal muscle and fat grown from cells in a tank, and here I want to be careful rather than hopeful, because the real picture is mixed and the field’s own press releases are not a reliable guide to it. The progress is real. Cultivated meat is now legally sellable in Singapore, the United States, Israel and Australia, and the cost has fallen by orders of magnitude, from around 325,000 dollars for the first lab-grown burger in 2013 to a growth medium that one formulation brought down to 0.63 dollars a litre.10 Unfortunately, the trouble is everything around that number. This sector only raised about 74 million dollars in 2025, down from 1.3 billion in 2021, and remains almost entirely pre-revenue.
Even worse, seven American states have now banned these products outright.11 And the deepest objection is not political but physical. David Humbird’s analysis, commissioned by people who wanted the technology to succeed, argued that the biology imposes a cost floor around 21 dollars a kilo that may keep cultivated meat permanently above commodity chicken, and a 2024 review of all the published studies landed in the cautious middle, that parity is unlikely without advances no one has made yet.12 So I am not promising you a cheap cultured chicken breast next year, or maybe next decade. I am pointing at the direction of the curve, and at the fact that the nearer bet, fermentation, is already on the shelf.
What makes me think the curve gets followed even through the bans and the funding winter is that the animal is not the only thing these technologies make unnecessary, and the other things they remove are things the market already wants gone. Most of the world’s antibiotics are not used on people. They are fed to farm animals, which is the single largest driver of the resistance that, on current trends, will kill more people than cancer.13 Protein grown in clean tanks does not need them. It does not incubate the next influenza the way a way overcrowded shed for 100,000 birds does. It uses a fraction of the land, water and feed inputs. If those efficiencies survive scale-up, they create a path to lower costs. A government that bans cultivated meat to protect its ranchers is not protecting them from a fad. It is standing in front of a falling cost, and I believe that is not a thing governments manage to do for very long. The serious money has already seen which way it runs: China alone put something like 555 million dollars into biomanufacturing in a single year.14
None of this is guaranteed, and the timeline seems way too long, still. The setbacks are real, the lobbied bans are spreading in the US, and the most optimistic parity dates come from the people with the most to gain from our optimism. But the shape of the thing is pointing the right way, and it is pointing there for reasons that have nothing to do with whether anyone’s heart is in the right place. The market does not have to care about the chicken to stop needing the chicken. That is the whole point, and I see it as a much better foundation than care, because care is scarce and uneven and easily distracted. A price advantage just compounds, quarter after quarter, in every country at once, and it works as hard on the person who never thinks about animals as on the person who cannot stop.
I find I am not troubled that it will happen this way: that the largest moral improvement of the century will be carried out mostly by people trying to build a cheaper nugget and book a better quarter, and that the vegans who were right for 50 years will get none of the credit. If the killing stops, it will not matter to a single animal spared whether it stopped because we finally felt the right thing or because the right thing finally went on sale. It will only matter that it stopped.
Discussion on HackerNewsFaunalytics, Global Animal Slaughter Statistics & Charts, analysis of FAOSTAT (“Producing Animals / Slaughtered”). ↩︎
Our World in Data, How many animals are factory-farmed?, from FAOSTAT. ↩︎
The FAO reports fish by weight, not by number. The per-animal figures are tonnage divided by an estimated mean weight per species, which is why the ranges are so wide. ↩︎
Mood & Brooke, Animal Welfare 33 (2024): e6 (1.1–2.2 trillion wild finfish/yr); Mood et al., Animal Welfare (2023) (~124 billion farmed finfish); fishcount.org.uk (255–605 billion farmed crustaceans). ↩︎
Rethink Priorities, 1–1.2 trillion insects farmed per year. ↩︎
An estimated 6.5 billion male chicks are culled each year, soon after hatching. Germany and France have banned the practice. ↩︎
The UK’s Animal Welfare (Sentience) Act 2022 was extended to decapod crustaceans and cephalopods after an LSE review of the evidence. ↩︎
OECD-FAO Agricultural Outlook 2025-2034: ~365 Mt of meat in 2024, poultry ~43% of meat protein by 2033; aquaculture overtook wild capture in 2022. ↩︎
Perfect Day (2020) and Remilk (2022) hold US GRAS status for fermentation-made dairy proteins, sold in the US and Singapore. ↩︎
Cultivated meat is sellable in Singapore, the US, Israel and Australia. The first cultured burger cost ~$325,000 (2013); a Hebrew University / Believer Meats serum-free medium reached $0.63/litre. ↩︎
GFI, State of the Industry: cultivated-meat firms raised ~$74 million in 2025, down from $1.3 billion in 2021. Seven US states had bans in effect by end of 2025. ↩︎
Humbird, Biotechnology and Bioengineering (2021): a cost floor around $21/kg. A 2024 Nature Food review found parity with conventional meat unlikely without major advances. ↩︎
Van Boeckel et al., PNAS (2015): most antimicrobials by volume go to farmed animals, not people. The 2016 O’Neill Review projected up to 10 million AMR deaths a year by 2050. ↩︎
China committed ~$555 million to biomanufacturing in 2025. ↩︎