I mean, AI researchers when I was in college laughed at using any kind of neural network.
And they were right, if you had a short timeframe.
I do think the timeframes are, uh… optimistic like a new SPAC with no revenue.
Overly optimistic.
No, it's not. That's what the whole article is about. It won't be cost effective, not even at that scale. Plus, it gives some other physical/pragmatic reasons on why it may not be even possible to do it (disregarding costs).
However the reports it mentions are pretty good:
GFI (Good Food Institute which represents these lab meat companies) https://gfi.org/wp-content/uploads/2021/03/cultured-meat-LCA...
Ezra Klein/Vox/Huw Hughes: https://www.scribd.com/document/526220188/Cultivated-Meat-re...
Open Philanthropy/David Humbird: https://engrxiv.org/795su
No one is ignoring it. It is already baked into the cost of the meat.
However, I also read an article [3] that seems to indicate that the complete opposite was true as farmers who use BLM land have to maintain it which made public land cost $1.20 more than private (2011). However, I don't really trust this that much as the guy who wrote it was a farmer and on the "Public Lands Council" which is a pro ranching organization. [4]
[1] https://www.blm.gov/press-release/blm-and-forest-service-ann... [2] https://sgp.fas.org/crs/misc/RS21232.pdf [3] https://thehill.com/blogs/congress-blog/economy-budget/27344... [4] https://publiclandscouncil.org/
(I didn't spent a lot of time researching but I found those articles interesting so hopefully you will as well!)
Are you contending that most beef interacts with BLM properties in some fashion?
The third link is just a counter argument to that says although the land looks cheap in reality its more expensive because of land maintenance.
However, I don't trust that article a lot as it was written by a rancher that is in a pro ranching organization so there is a big bias.
While they are a subsidy, I do not believe they are relevant at all to the argument. That's my point, I guess.
Corn, bean, and other row crop subsidies, sure. But BLM is just a distraction, I think.
Edit: The $38bn appears to also include egg and dairy subsidies, but Americans appear to spend substantially less per annum on those commodities compared to meat[3][4], such that the federal subsidies still appear to make up a rather large fraction of the (retail value of!) annual consumption.
[1]https://meatonomics.files.wordpress.com/2013/06/sample-chapt...
[2]http://www.themarketworks.org/stats
[3]https://www.statista.com/statistics/417446/us-dairy-departme...
And some of these crops are used in finishing lots.
> but cattle are no more subsidized than housing or roads or cars or computers or any of the other things that make life better.
I think it’s questionable that cattle make “like better”. They are an inefficient form of food who suffer when we kill them, and we don’t need to eat them to eat delicious, nutritious food.
A cow walks around and eats grass by itself. A bioreactor needs constant monitoring by a worker.
If each and every beef cow was hand-raised by a single human, regular beef would be $100/kg too.
But even that’s really just the first attempt, breweries are a reasonable comparison long term and batch sizes get huge.
You might enjoy the Humbird report: https://engrxiv.org/795su
Muscle cells can't tolerate shear forces yeast can, are much more easily poisoned by their own wastes, have a doubling time 12+ times slower, require sterile pharmaceutical-grade amino acids as a feed stock (which themselves cost $100+/kg), etc etc.
Humbird says commercial cell culture bioreactors don't get any bigger than 25 m^3, and have to be run in a clean room. Doesn't sound much like any brewery I've seen.
With plenty of investment you might get 2x better on any one of those metrics, but in order to compete with cows you have to be 100-1000x better on every single one.
They address one of your criticisms right at the beginning: “The replacement of amino acid media with plant protein hydrolysates is discussed and requires further study.”
It’s a solid analysis in the short term, but optimizing cell lines can significantly change some of these numbers.
it will certainly require some training to attend to a bioreactor but assuming an eventual working model they seem much less problematic than a wandering animal, and would probably suffer less and get sick less than lots of caged animals.
> you can't leave cows unattended.
Of course you can. There are remote areas that essentially let cattle roam freely over thousands and thousands of acres and bring them in as little as once a year for maintenance (branding, medication, castration, selling, etc). No fences, no roads, rarely ever being seen aside from routine checks of water points (which is becoming less common with remote monitoring equipment).
> they wander off, get injured, fall ill, might be stolen, and all kinds of things. if they don't eat well enough they might not be worth slaughtering.
> you can either keep them in cages and attend to them there, with all the distasteful complications that brings, or herd them in large pastures at the cost of very difficult and dangerous skilled labor.
There is not much dangerous skilled labor at all. Difficult, yes like a lot of manual labor. Dangerous and skilled no outside of some rare cases like the use of helicopters to drive cattle in aforementioned remote areas with terrain sometimes not even accessible to motorbike so helicopter and horse is used. Importantly even in those situations, the amount of labor required per value of beef produced is simply not very high.
And those aren't really the options. Feedlots are increasingly used (what distasteful complications?) but most of the world's beef is not produced that way, and where those are used it is often as a finishing step after a free-range life to improve consistency and make up the last bit of weight.
There is very little that is difficult and dangerous about the labour required to herd cattle a couple of times a year, unless you are talking about the gyrocopter cowboys in far north queensland and the Northern Territory, of whom a better description would be crazy or insane
Feedlotting is more labour intensive, but then the beef is more expensive because in some markets it is considered of a higher grade, and they may weigh more
Humbird estimated $17-$40 for a pound of ground meat by 2030. Future Meat has already reached $18 for cultured ground "chicken" but as the article points out is unlikely to go lower. Currently, ground beef costs about $4 per pound. $4. That's how much everything, from infrastructure to land use to feed to water to labour to delivery costs for the same ground beef you'd get from livestock as you'd get straight out of the bioreactor (which they said was 70% water and 30% basically ground meat). A round steak costs about $6 per pound if you're lucky, and often higher. Going for steak with meat cultures this would drive the price well beyond $17 since you'd need to do a LOT more work in processing it.
Livestock has been scaled out very efficiently by ignoring all the environmental damages it causes. Since the whole point of this is to not "cut corners", they're already on the back foot, and the article explains many of the concerning avenues they are considering to even try and get close to dropping the price.
Ultimately, livestock has been optimised, and isn't getting cheaper, quite the opposite. Prices are slowly rising for meat. But the "hidden" cost is the environmental damage (a damning cost at that). Scaling out the bioreactors physically with higher capacity is how you'd drop the cost of cultured meat at the same volume, fewer larger reactors being more efficient than many smaller ones, since the pharmaceutical industry has failed to make game changing breakthroughs in the other approaches to making more efficient cultures that would meet this need in recent years (though, as mentioned, Future Meat has done just this, but would still need economy of scale to get competitive).
They could scale up tomorrow with ten thousand times as many reactors without making a real dent in their efficiency, simply replicating what they already have, and still be forced to sell at cost like Eat Just or sell at an unjustifiable price. Current cost according to the article is typically between $10k and $22k per pound. That's obscene. Getting down to $17 per pound, or even close such as Future Meat with it's incredible $18 per pound, is still a losing battle. After all, even if we removed subsidies and doubled the price of meat, that's still significantly cheaper than ground meat cultures, let alone being processed.
Meanwhile, plant-based alternatives and such are already reaching the point of being cost competitive, with $6 per pound for a Beyond Beef burger, and an average price of $9 per pound for meat alternatives, and some people are already switching. So if there's anything that governments should focus on, it's the one that's clearly successful (and in way more volume) and just about a similar win in terms of soy/grain usage and similar.
Is the article perfect? No. Is it correct? Most likely. I don't think the focus is off, though I do think that there's some ground work that could have been done, showing breakdowns of costs and such in a clearer way, though clearly it's prose in the end while the reports speak for themselves. Talking about the story of the analysis and what it actually means for the likely future of meat culture. Hopefully I've collected those points you were looking for a little bit here, though not being an industry expert naturally I've been unable to trivially procure breakdowns for how much is labour costs per pound etc., so mostly it's a denser version of the factoids in the article and what's readily available online.
There are a ton of subsidies provided by the government that are not accounted for in this cost. Water rights, land, delivery all are accounted for differently and I would not say with any measure of confidence that they are all represented well in the consumer cost. There is also the large externalities that are completely not accounted for like carbon emissions.
It is sort of like comparing electric cars vs gas back when people would comment how cheap oil is and ignore the wars fought over it and the environmental devastation wrought from its production.
In the end, if the seller wants the same margins as before, then they'll raise by exactly as much as the subsidy covered. Even if it's the higher amount and a bit over double the price, that's perhaps $10 for a pound of ground beef. That's practically half the price of cultured meat, and about the price of meat alternatives; meat alternatives being currently a high margin business as companies like Beyond Meat and Impossible Foods are trying to increase R&D and production to meet demand, so they could actually drop prices now, they just won't since they have nowhere near the supply.
Adding similar subsidies to meat culture products would simply narrow the gap, fundamentally they are more expensive and that's even with the "optimistic" estimates Humbird gave. Reminder that Future Meat is an impressive $18 per pound for chicken cultures, whereas actual chicken runs about $2 per pound. Meanwhile, meat alternatives around $9 per pound TODAY will improve by 2030 to be much more cost competitive than meat cultures seemingly will be able to, unless they can manage all these frankly miraculous breakthroughs and get as cheap as chicken (based on the $5.66 per kilo figure, or little over $2 per pound), which seems unlikely.
As I saw it put nicely elsewhere, removing the subsidy would result in a "more accurate, divorced-from-government cost of their inputs and externalities which affect the health of everyone, to include the consumer and the planet at large."
Please enumerate at least 5. You can't because there are not a "ton" of subsidies for meat production unless.
That number is the count of full grown cattle, 31M. When you include calves, it's 98M total in the US. It isn't easy to realize just how nauseating that truly is.
There's roughly 330M people in the US. One cow will feed 2300 people 3oz of beef, which is the most meat one may eat in a sitting and still be considered healthy (though I personally disagree, no amount of red meat is healthy IMO). 360M / 31M comes out to around 10 people per cow, which will feed them once a day 3oz for 230 days. So the US has enough beef to serve every man, woman and child a a 3oz steak serving once every 38 hours, all year long.
Of course, nothing close to this amount is actually consumed, and at least 23% of it is entirely wasted in production, 7.1M cattle annually are effectively turned into garbage before product can hit store shelves.
I’m sure I eat, on average, at least 6 oz/ day. I estimate I eat anywhere between 50-100 kg/ year or somewhere between 5-10 oz a day. Probably closer to around 7. Or about three times your estimate.
Also, I didn't decide anything. Science decided that all those cows are the source of upwards of 4% of US greenhouse emissions, and science also decided that eating beef massively increases risk of type 2 diabetes, coronary heart disease, stroke, colon cancer, and, uh, oh yeah, death. And though there would be no greenhouse emissions nor any health risk to you or others in so doing, I have decided neither you nor anyone else anywhere should eat my toenail clippings.
The part I dislike was focusing a lot on the size of the reactors needed in order to replace the US meat production. It is a bit unfair to compare the size of a bioreactor to a meat packing plant without considering the size of all the (organic?) bioreactors... livestock. There are ~30M cattle in the US at any given time so the volume is considerable. To ignore the infrastructure and space that livestock use is making the implicit claim that livestock are 'free' instead of 'cheaper currently'.
Considering how efficient meat production is (at the expense of everything else), we can basically make a single head of cattle as our standardized unit to evaluate the cost & thermodynamics.
Basically at scale is a engineered bioreactor cheaper than the natural blood & bones bioreactor? Certainly not upfront unless small scale bioreactor in a bag technology... oh wait that exists [1]. Sucks that pricing isn't easy to get but I think it is still more expensive than fixed/reusable vessels at least over time.
[1] https://www.sartorius.com/en/products/fermentation-bioreacto...
It's really hard to know without having more knowledge about the specifics of but I think there's a lot of unexplored areas that makes these projects worthwhile.