Daniel sent us this one — and it starts with a bottle. A bottle of milk sitting on a shelf in Jerusalem that never touched a cow. The brand is Remilk, it's got a photo of an elderly gentleman in one of those flat caps that says "trustworthy farmer" on the label, and it's certified parve. Which means, in kosher terms, neither milk nor meat. And yet it's milk. Or is it? That question — what do we call a thing that is chemically identical to something it biologically is not — turns out to be one of the most interesting questions in both food science and religious law right now.
It's not hypothetical anymore. Remilk got its kosher parve certification from the Israeli Chief Rabbinate in twenty twenty-three. As of now, mid twenty twenty-six, you can walk into stores in Israel, the US, and Singapore and buy precision-fermented dairy from at least four different companies. This is on shelves. The future of food arrived while we were all arguing about oat milk.
Which, by the way, is not what this is. I think that's the first thing to clear up. Plant-based milks — oat, almond, soy — those are emulsions of ground-up plant material in water. Completely different protein profiles. This stuff contains actual whey protein. Beta-lactoglobulin, specifically. The dominant whey protein in cow's milk. It's just that no cow was involved in making it.
So let's open that bottle and look at what's actually inside — both the science and the legal framework that surrounds it. Because what's in there is a product of something called precision fermentation, and it represents a genuine category crisis for how we regulate food, certify food, and even think about food.
I like that. The milk that broke the spreadsheet.
So precision fermentation — let me define this properly. You take a microorganism, typically a yeast or a fungus, and you genetically engineer it to produce a specific protein. The organism becomes a tiny bioreactor. You feed it sugar, it grows, and it pumps out the protein you want. Then you filter out the organism, purify the protein, and you've got your ingredient. In Remilk's case, the protein is beta-lactoglobulin, and the microorganism is a yeast called Komagataella phaffii. It used to be called Pichia pastoris — same organism that Impossible Foods uses to make their heme protein for the Impossible Burger, by the way.
It's the same yeast, different genetic instructions. One makes a blood-tasting protein for fake meat, the other makes milk protein for fake milk. Same factory, different product.
Same factory, different product — that's a great way to put it. And the process is actually elegant. Step one: you synthesize the gene for beta-lactoglobulin. Step two: you insert that gene into a plasmid, which is basically a little ring of DNA that the yeast will accept. Step three: you introduce that plasmid into the yeast cells — this is called transformation. Step four: you grow the transformed yeast in stainless steel fermenters with a glucose feedstock. Remilk uses a proprietary yeast strain that secretes the protein directly into the growth medium, which simplifies the downstream processing enormously. Instead of having to break open the yeast cells to get the protein out, you just harvest the liquid.
The yeast sweats milk protein into the broth.
Then step five: you filter and purify the protein from the broth using filtration and chromatography. Remilk claims about ninety percent recovery efficiency in purification. Step six: you formulate the final product — you add fats, vitamins, minerals, and water to create something that looks and tastes like milk. Because what comes out of the fermenter is just the protein fraction. It's not milk yet. It's a protein ingredient.
This is where the nutritional claims get interesting. The bottle Daniel mentioned says seventy-five percent less sugar and three percent fat. That sounds impressive until you realize what's actually going on. The sugar in regular milk is lactose. Precision-fermented milk has no lactose because the fermentation process only produces the protein — not the full milk matrix. So "seventy-five percent less sugar" really means "we didn't add the sugar back in." It's not that they removed it. It was never there.
And the three percent fat? That's added separately, typically from plant oils — sunflower oil, coconut oil. So the fat profile is completely different from bovine milk, even though the protein is identical. You're getting cow-identical whey protein swimming in a plant-based fat emulsion. It's a chimera of a beverage.
A chimera of a beverage. The glockenspiel of corporate approachability.
I'm not even sure what that means, but I'll allow it. Here's the thing about the economics, because this is where the rubber meets the road. Remilk's facility in Rehovot, Israel, operates twenty-thousand-liter fermenters. A single batch runs five to seven days. And Remilk claims that one liter of fermentation broth yields the protein equivalent of fifty liters of cow's milk. So one twenty-thousand-liter fermenter, running one batch, produces the protein equivalent of roughly a thousand cows per day. That's staggering.
A thousand cows don't require stainless steel fermenters, glucose feedstock, chromatography columns, and a team of bioprocess engineers.
The capital costs are enormous. Current estimates put precision-fermented whey protein at five to eight dollars per kilogram of protein, compared to one to two dollars for conventional whey. That gap is narrowing — it was more like ten to one five years ago — but it's still substantial. The companies are betting that scale will close it further, and that consumers will pay a premium for animal-free dairy. Perfect Day, the US-based company that's been doing this since twenty twenty, now supplies whey protein to over fifty brands — ice cream, protein powders, cream cheese. They've proven there's a market.
The science is elegant — yeast doing what yeast do, just with a new set of instructions. But the question of what to call the result? That's where things get complicated, and it's a complication that's been brewing for centuries in Jewish legal tradition.
Let me set up the kosher paradox here, because it's genuinely fascinating. In Jewish dietary law, milk and meat cannot be mixed. Not just in the same dish — you can't eat them in the same meal, you need separate dishes, separate utensils, separate dishwashers in many homes. The separation is absolute. And the category "dairy" is defined by origin: if it comes from a kosher mammal, it's dairy. So what happens when the protein is chemically identical to cow's milk protein but came from yeast?
You've got a substance that looks like milk, tastes like milk, behaves like milk in cooking, has the same protein structure as milk, but was produced by a microorganism that is itself neither milk nor meat. Parve is the default category for things that are neither. Plants are parve. Yeast is parve. Water is parve. So if yeast makes a protein, and yeast is parve, is the protein parve?
This is where the responsa literature gets rich. Since about twenty twenty, there have been three major positions emerging among Orthodox poskim — Jewish legal decisors. And I should say, as Daniel noted in the prompt, nothing we say here is rabbinic guidance. Consult your local rabbi. We're describing a debate, not resolving it.
So walk me through the three positions.
Position one: the molecular origin view. This is associated with Rabbi Yisrael Belsky, of blessed memory, and Rabbi Hershel Schachter. The argument is that since the gene used to program the yeast came from a cow, the resulting protein inherits the dairy status of the gene donor. The genetic information originates in a dairy animal, therefore the product is dairy. It's a kind of informational lineage argument — the protein "remembers" where its blueprint came from.
That's a fascinating way to think about it. The gene as a kind of legal chain of custody. The protein is dairy not because of what it is, but because of where its instructions came from.
And Rabbi Belsky actually addressed a parallel case in twenty nineteen with Impossible Foods' heme protein. He ruled that the heme — which is soy leghemoglobin produced in Pichia pastoris yeast — was parve, because the gene came from soy, which is parve. So by that logic, if the gene donor is parve, the product is parve. If the gene donor is dairy, the product is dairy.
Under that framework, Remilk would be dairy, because the beta-lactoglobulin gene came from a cow. But wait — the Israeli Chief Rabbinate certified Remilk as parve. So they're clearly not following position one.
Which brings us to position two: the process-based view. This is advanced by Rabbi David Stav and the Tzohar rabbinate in Israel. The argument is straightforward: no animal milk is involved at any stage of production. The yeast is a parve organism. The feedstock is plant-based glucose. The final product has never been inside a cow. Therefore it's parve, regardless of where the gene came from. What matters is the actual production process, not the informational history.
This feels like the difference between asking "what is this made of right now" versus "what is the genealogy of its components." One is chemistry, the other is biography.
That's beautifully put. And the key halakhic precedent here is rennet in cheese-making. Traditional kosher cheese requires animal rennet from a kosher-slaughtered calf — rennet is an enzyme from the stomach lining that coagulates milk. But in the nineteen sixties, microbial rennet produced by fungi was accepted as parve by most authorities. The reasoning: the fungus is not an animal, the enzyme it produces is not an animal product, therefore the cheese made with it doesn't inherit any meat or dairy status from the production process. This is the closest analogy to precision-fermented milk, and it's cited extensively in the responsa.
If microbial rennet is parve, microbial whey should be parve.
That's the argument. And the Chief Rabbinate accepted it for Remilk. But there's a third position, and this one is more about social policy than molecular biology. Some Haredi poskim — including Rabbi Moshe Sternbuch, head of the Edah HaChareidis rabbinical court in Jerusalem, who issued a ruling in twenty twenty-four — argue that even if the product is technically parve, it should be treated as dairy. The reason is something called mar'it ayin — the appearance of impropriety.
Explain mar'it ayin.
It's a principle in Jewish law that you should avoid actions that, while technically permitted, could lead an observer to mistakenly think something forbidden is happening. If you sit down in a restaurant and drink what looks exactly like milk while eating a steak, someone watching might think dairy-meat mixing is permitted. So even if the substance is parve, consuming it with meat could create confusion and erode the broader observance of the law.
It's not about what the product is. It's about what it looks like. The legal status of the thing is less important than the social signal it sends.
And this isn't a fringe concern — mar'it ayin has been a serious halakhic principle for centuries. It's why some kosher households don't use almond milk in coffee served with a meat meal unless the almond milk container is left on the table, so people can see it's not dairy. The visual evidence matters.
You've got three positions. Position one: it's dairy because the gene came from a cow. Position two: it's parve because no cow was involved in production. Position three: it might technically be parve, but you should treat it as dairy anyway to avoid confusion. Three different answers, all from serious scholars working with the same texts and the same facts.
There's a fourth wrinkle I want to add, because it's elegant. Rabbi Asher Weiss, one of the most prominent poskim in Israel today, issued a responsum in twenty twenty-three ruling that precision-fermented milk is parve. But his reasoning is different from the process-based view. He draws an analogy to honey.
A bee is not a kosher insect. You can't eat a bee. But honey is kosher. Because honey is not a product of the bee's body — it's a secretion. The bee collects nectar, processes it, and deposits it. The honey never was part of the bee. Rabbi Weiss argues that the whey protein secreted by yeast is analogous: the yeast produces it and releases it into the medium, but the protein was never part of the yeast's bodily structure. It's a secretion, not a body part. Therefore it's parve.
That's gorgeous reasoning. It completely sidesteps the gene-origin question by focusing on the physical relationship between the organism and the product. The bee is treif, the honey is kosher. The yeast is parve, the whey is parve. The genetic engineering is irrelevant to the category analysis.
This is what I find so compelling about this whole debate. These are serious legal minds applying frameworks developed over two millennia to a technology that didn't exist five years ago. They're not just making things up. They're reasoning by analogy, by principle, by precedent. The same way secular judges handle new technologies in contract law or intellectual property.
What strikes me is that none of these positions is obviously wrong. Each one is internally coherent. The disagreement isn't about facts — everyone agrees on what precision fermentation is and what it produces. The disagreement is about which legal principle should govern. Origin, process, or appearance. And halakha doesn't have a single meta-rule for deciding which principle takes priority when they conflict.
And that's why there's no consensus, and probably won't be for a while. The Israeli Chief Rabbinate certifies Remilk as parve. The OU in the United States has taken different positions on different products — they certified Perfect Day's whey as dairy for some applications and parve for others, depending on how the product is formulated and marketed. Some Haredi consumers simply won't touch any of it regardless of certification. The market is fragmenting along communal lines.
Which brings us to the practical question. If you're a kosher consumer standing in the supermarket holding this bottle, what do you do?
You look at the specific certification on the package and you consult your local rabbi. There is no single answer. The OU, Star-K, the Israeli Chief Rabbinate, Badatz, Rabbi Landau's certification — they have different policies. And even within the same certification agency, different products may receive different rulings depending on how they're produced and marketed. A precision-fermented whey powder sold as an industrial ingredient might get one ruling, while a ready-to-drink milk sold in a dairy-style carton might get another.
The consumer needs to be a kind of amateur posek themselves. Or at least know which posek they're following.
Which, honestly, has always been true in kosher observance. Different communities follow different authorities. This is just a new domain where the differences become visible.
What's striking about this debate is that it's not really about milk. It's about how any legal or regulatory system handles something that is simultaneously new and identical to something old. The same question is coming for lab-grown meat, for synthetic eggs, for precision-fermented collagen and gelatin. The frameworks we build now for milk will set precedents for all of those.
The lab-grown meat case is actually harder, in some ways. With milk, you're dealing with a secreted protein. With cultured meat, you're taking actual animal cells — stem cells from a cow or a chicken — and growing them in a bioreactor to produce muscle tissue. The cells themselves are animal cells. They have the DNA of the donor animal. So the halakhic questions multiply. If the original cells came from a living animal without slaughter, is that a form of ever min ha-chai — eating a limb from a living animal, which is prohibited? If the cells came from a kosher-slaughtered animal, does the slaughter count if no one was slaughtering the animal for food at the time? And if the cells are grown on a scaffold made of non-kosher materials, does that affect the status?
It's like the milk debate but with every dimension turned up. Origin, process, appearance, and now you add the question of whether lab-grown tissue even counts as "meat" in the halakhic sense. Because meat comes from an animal that was slaughtered. If there's no animal and no slaughter, is it meat?
There's a fascinating responsum from Rabbi Shlomo Aviner on this. He argues that cultured meat is parve regardless of the cell source, because the definition of meat in halakha requires an animal body. No body, no meat. The cells are just cells. Others disagree strenuously. Rabbi Asher Weiss, who gave us the honey analogy for milk, has suggested that cultured meat might be meat because the cells are identical to animal tissue and the product is functionally indistinguishable from slaughtered meat. Same rabbi, different product, different reasoning — which tells you how case-specific this all is.
This is where the secular parallel becomes really clear. The FDA and the USDA spent years fighting over who gets to regulate lab-grown meat. Is it the USDA because it's meat, or the FDA because it's produced through biotechnology like a pharmaceutical? They eventually reached a joint regulatory framework in twenty nineteen, but it took years of turf warfare. The same category problem, just with agencies instead of poskim.
The FDA's "generally recognized as safe" process is basically secular kosher certification. It's a bunch of experts reviewing a novel food and deciding whether it fits into existing categories or needs new ones. The questions are almost identical: what is this stuff, how was it made, is it substantially equivalent to something we already know is safe, and what do we call it on the label?
The labeling wars are their own thing. The dairy industry has been fighting to prevent plant-based and precision-fermented products from using the word "milk" on the label. The argument is that milk is defined by its origin — it comes from a mammal — not by its nutritional profile.
It's exactly the same debate. Origin versus composition. The dairy lobby is basically taking the molecular origin position, while the alt-protein companies are taking the process-based position. The secular world is replaying the halakhic argument with lobbyists and federal judges instead of rabbis and Talmudic precedents.
Where does this leave us practically? Let me try to pull some threads together. For kosher consumers, precision-fermented dairy is a genuine innovation that solves a real problem. If it's parve, you can have a cheeseburger that tastes like a cheeseburger but is fully kosher. You can put milk in your coffee after a meat meal. You can have ice cream for dessert without waiting six hours. The potential is enormous.
For the food industry, precision fermentation is not a single product — it's a platform. The same yeast strains and fermenters that produce milk proteins can be reprogrammed to produce egg whites, which is what Clara Foods is doing. Or collagen, which is what Geltor is doing. Or even human breast milk proteins, which is what TurtleTree is working on. The regulatory and religious frameworks being established now will set precedents for all of these.
There's an Israeli company called Imagindairy, founded by a former Remilk scientist, that's developing a process to produce the full milk fat profile — not just the whey protein, but the actual milk fats that give dairy its mouthfeel and flavor complexity. If they succeed, the product will be molecularly indistinguishable from cow's milk in every component. Protein, fat, sugar — all identical, none from a cow. At that point, the "is it really milk" question becomes almost philosophical.
The halakhic question becomes even more acute. If every molecule is identical to cow's milk, but no cow was involved, what is it? At some point the law has to decide whether "milkness" is a physical property or a historical fact. You can't measure history with a mass spectrometer.
A mass spectrometer can tell you what a thing is. It can't tell you where it came from or what it means. And meaning is what law — religious or secular — is supposed to provide.
Let me add one more dimension that I think is underappreciated. There's a potential future where kosher-observant Jews become early adopters of precision-fermented dairy specifically because it solves the meat-milk separation problem. If the product is widely accepted as parve, it opens up entire categories of cuisine that were previously off-limits. That's a huge market incentive for kosher consumers to push for the permissive rulings.
The flip side is also possible. If the halakhic uncertainty persists, kosher consumers might become late adopters, waiting for a consensus that never quite arrives. And the Haredi community, which tends to be more stringent on new food technologies, might simply opt out entirely. So you could get a bifurcation — Modern Orthodox homes using precision-fermented parve milk in their coffee after meat, while Haredi homes treat it as dairy or avoid it altogether.
Which is already happening. Remilk is on shelves in Israel, and different communities are making different choices. The market is sorting itself out along the same lines as the responsa literature.
Let me zoom out for a second, because I think there's something bigger here. What we're watching is the process by which a society develops norms around a novel technology. It's messy. It involves competing authorities, analogical reasoning from old cases, and arguments about appearances versus reality. This is how it always works. The rabbis debating precision fermentation are doing exactly what the rabbis did with electricity on Shabbat, with artificial insemination, with organ transplantation. They're applying ancient principles to new facts, and reasonable people disagree about which principles apply.
The secular world does the same thing. The European Union's novel foods regulation, the FDA's GRAS process, the labeling fights — it's all the same dynamic. A new thing appears, it doesn't fit the existing categories, and we have to decide whether to stretch the categories or create new ones. The kosher debate is just a particularly well-documented and intellectually rigorous version of a universal problem.
There's one more precedent I want to mention, because it's illuminating. Gelatin is made from animal collagen, usually from pigs or cows. For decades, there was a major halakhic debate about whether gelatin is kosher. The argument for permissiveness: the manufacturing process chemically transforms the collagen so thoroughly that it's no longer considered food derived from an animal — it's a new substance. The argument against: it comes from a non-kosher animal, and the transformation doesn't change its essential identity. Different authorities ruled differently.
It's the exact same structure of debate. And gelatin was eventually accepted as kosher by many authorities for certain applications, though not all. The OU certifies gelatin from kosher-slaughtered animals, and some authorities accept gelatin from non-kosher animals if it's been fully chemically transformed. But there's still no universal consensus, and some kosher consumers avoid all gelatin unless it's fish-derived or plant-based.
The gelatin debate took decades and still isn't fully settled. The precision fermentation debate is maybe five years old. We're at the very beginning of this.
The technology is moving faster than the law. By the time a broad halakhic consensus emerges on precision-fermented whey, the companies will have moved on to precision-fermented casein, or full-spectrum milk fats, or hybrid products that combine fermentation-derived proteins with plant-based fats and cultured animal cells. The poskim are going to be playing catch-up for a generation.
Which is not necessarily a problem. Jewish law has always developed through case-by-case adjudication rather than top-down legislation. The responsa literature is basically a twenty-five-hundred-year conversation. This is just the latest chapter.
It's a chapter that has real consequences for millions of people. The global kosher market is estimated at over three hundred billion dollars annually. Kosher certification affects what products can be sold in Israel, what ingredients food manufacturers use, and what observant Jews around the world can eat. When a posek rules that precision-fermented milk is parve or dairy, that ruling ripples through supply chains and supermarket shelves.
Let's bring this back to the bottle Daniel was holding. Remilk, sitting on a shelf in Jerusalem, certified parve by the Israeli Chief Rabbinate. It's a physical object that contains a legal argument. Every bottle is a claim: this is not dairy, regardless of what your senses tell you. And whether you accept that claim depends on which rabbi you follow, which community you belong to, and which legal theory you find more compelling.
The bottle also contains a scientific claim: we can produce the same proteins as a cow, more efficiently, without the cow. The twenty-thousand-liter fermenter in Rehovot making the protein equivalent of a thousand cows per day — that's an argument about the future of food that goes far beyond kosher certification. It's about land use, water use, methane emissions, animal welfare. The kosher debate is happening inside a much larger conversation about whether precision fermentation can actually compete with factory farming.
On the economics, the Good Food Institute published a report in twenty twenty-five valuing the global precision fermentation market at about two point two billion dollars, projecting it to reach eight point five billion by twenty thirty. Those are big numbers, but they're still tiny compared to the global dairy market, which is over eight hundred billion. Precision fermentation is real, it's growing, and it's still a rounding error in the dairy industry.
The cost curve is the thing to watch. Five to eight dollars per kilogram of protein versus one to two dollars for conventional whey — that gap has to close for this to be more than a premium niche product. The companies argue it will close as they scale, as the technology improves, as feedstock costs come down. The skeptics point out that conventional dairy is also getting more efficient, and that cows have a few hundred million years of evolutionary optimization on their side.
Cows are basically self-replicating, solar-powered, grass-fueled bioreactors. They're hard to beat on capex.
They do have certain advantages, yes. Though the methane problem is real, and if carbon pricing ever gets serious, the economics shift.
Let me ask you something. You're a retired pediatrician. You've spent a career thinking about nutrition. If a patient asked you whether to give their kids precision-fermented milk, what would you say?
I'd say it depends on what you're comparing it to. Compared to cow's milk, the protein is identical, but the fat profile is different because it's plant oils instead of milk fat. For most kids, that's probably fine — the plant oils are typically chosen to have a reasonable fatty acid profile. The lack of lactose might be a benefit for kids who are lactose intolerant, which is most of the world's population. But you're also missing some of the minor components of milk that may have biological effects — immunoglobulins, growth factors, microRNA. We don't fully understand what all of those do.
It's nutritionally similar but not identical. The protein is a perfect match, everything else is an approximation.
And for adults using it as a milk alternative in coffee or cereal, the differences are probably negligible. For an infant or a small child whose primary nutrition comes from milk, I'd want to see more long-term data. But that's true of any novel food.
Which brings us back to the labeling question. If the nutritional profile is different, should it be called milk? If the protein is identical, should it not be called milk? The word "milk" is doing a lot of work here, and nobody agrees on what it means anymore.
The dairy industry says milk comes from mammals. The alt-protein industry says milk is defined by its function and composition. The rabbis are split. The regulators are split. The consumers are confused. It's a perfect storm of category collapse.
That's the phrase for this whole episode. When a category that's been stable for millennia — milk comes from animals — suddenly doesn't work anymore, and everyone has to scramble to build a new one.
The scrambling is happening in real time, in public, with real products on real shelves. That's what makes this moment different from hypothetical debates about future technologies. You can buy the bottle. You can drink it. You can decide for yourself what you think it is.
Let me leave us with a question that I think lingers beyond the specifics of milk and kosher certification. As precision fermentation scales and costs drop, and as the technology expands to egg proteins, collagen, gelatin, and eventually cultured meat, we're going to face this same question over and over: do we define food by its origin or its composition? And the answer, I suspect, is that we'll do both, inconsistently, depending on the product, the context, and who's asking. The kosher debate is a preview of a much larger societal negotiation that's just getting started.
The Imagindairy case is the one to watch. If they succeed in producing the full milk fat profile — not just the proteins, but the actual triglycerides that give dairy its sensory properties — then you'll have a product that is molecularly indistinguishable from cow's milk in every respect. At that point, the origin-versus-composition debate becomes almost metaphysical. If you can't tell the difference with any instrument, is there a difference?
There's a difference if you believe there's a difference. Which is, I suppose, the entire point of law.
If you're a kosher consumer, the practical takeaway is: check the certification, consult your rabbi, and know that reasonable authorities disagree. If you're not a kosher consumer, the takeaway is: watch this space, because the frameworks being built now for religious certification are going to shape how all of us think about and regulate lab-grown food for decades. The rabbis and the regulators are asking the same questions, just in different vocabularies.
If you're just someone who likes milk, the takeaway is: there's now milk that never met a cow, it's on shelves, and whether you think that's miraculous or unsettling probably says more about you than about the milk.
Now: Hilbert's daily fun fact.
Hilbert: In the eighteen forties, researchers studying cuttlefish ink discovered that the pigment sepia contains melanin granules that are so stable they can survive in fossilized cephalopod ink sacs for over one hundred sixty million years — meaning the ink from a Jurassic squid could, in theory, still be used to draw a picture.
The ink lasts longer than the squid. Good to know.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. If you've got a weird prompt about how ancient law meets synthetic biology, or anything else that sits at the intersection of two things that don't seem like they should intersect, send it to prompts at myweirdprompts dot com. You can find every episode at myweirdprompts dot com or wherever you get your podcasts. I'm Corn.
I'm Herman Poppleberry. Drink responsibly, whatever you decide is in the bottle.
