So Much Blood
dynomight.net
dynomight.net
"To get the actual data, you need to go through a website maintained by the US Trade Commission. This website has good and bad aspects. On the one hand, it’s slow and clunky and confusing and often randomly fails to deliver any results. On the other hand, when you re-submit, it clears your query and then blocks you for submitting too many requests, which is nice."Tangent, but: I wish Hacker News was a bit less dogmatic about preferring the original site's title over any better semantic summary the submitter might offer. Blog posts and news headlines have a strong incentive to be clickbait-ey or just plain catchy, which often makes them less informative and in turn makes the HN homepage harder to scan for interesting posts.
It also feels like an arbitrary dividing line where if the original title is too long to fit HN's max limit, the submitter edits it and their take on what's a good summary stands. But if the original title could fit, the submitter's headline is often overwritten with the article title by mods even if it's less useful.
I suppose I could create aliases and redirects instead.
The actual title of the page is "Mycoria".
I actually wanted to thank the submitter for extending the title to something meaningful, but thought that it would be a too low value-adding comment.
It was really easy, and I was surprised it apparently hadn't been done before.
The only problem is the index IP blocks you if you query it too often, and it gets queried once for every resource on the page.
I've been informed that there are ways around this (download an index?), does anyone know more about that?
Also, it's just a proxy, there's no front-end. (Being a 50 line Python script with one dependency (Flask), the idea is that the end user would run it.) It's just http://localhost/URL, and it loads it from the hardcoded snapshot. I guess https://localhost/snapshot/url might work better.
I'd also like to see how many lines it ends up being if you remove the one dependency. I've done this for other projects and the answer is "not much more" and "also runs way faster for some reason."
Second — it’s amazing the detail that you can achieve from public data.
Third — I’m left wondering if a true “Deep Research” like tool would be able to provide the same analysis. I find that Deep Research is fine for secondary sources, but not for Deep Analysis of primary source data.
Source: A guest lecture at my university by Al Roth, Nobel prize-winner in economics, who is currently focusing his work on these type of markets. Most of his work is on kidney exchanges right now.
"The chickens are harvest when they’re 32 days old"
Let’s sprout some semence in the cow (or not).
We only need a way to harness the power of the human body. Maybe we put people in VR for fun while using their body heat to power the AI.
That's explosives are chemicals selected / designed to be able to release their chemical energy really quickly and without needing any external oxidizer (because harvesting atmospheric oxygen would be too slow). That focus obviously leads to compromises in other areas, like energy density.
Also no human is anywhere close to being as knowledgeable and skilled as LLMs at all the things at the same time, so it hardly even compares.
lol, the spoiled times we live in that you think this. The human body is capable of surviving on very little.
A thing with protein, fat and sugar would sustain you for incredible amounts of time. Many many months if not years.
Human bodies have evolved to survive for a long time on relatively little, yes. But not to evolve for a very long time on a single source of very 'clean' food like snickers bars. 'Clean' in the sense that chemically snickers has relatively well defined inputs, whereas hungry humans would eat just about anything, including insects and grass and bark or leather.
I’m not talking about what it takes to stay alive for long term periods. I was refuting the silly idea that you would die after a couple of days/weeks of snickers.
How much vitamin A? Night blindness. Vitamin B? Neurological issues, confusion.
That's the thing with mono-diets, your body needs a diverse range of things that it can't synthesise itself.
But to the core point, in cases where the output of an LLM is good enough, many already have much lower energy requirements than humans: o4-mini is currently priced at $1.1 per million tokens of input and $4.4/million tokens of output; if that's all being spent on electricity at $0.1/kWh, that's a max of 11 kWh/million tokens in and 44 kWh/million tokens out — how many calories would a human have to burn to read, write, hear, speak, and internally monologue the equivalent of a million tokens?
I wanted to say that you were wrong, that LLMs can't reason and so it certainly isn't an obvious truth that they do it better than humans, but when I asked AI if LLMs can reason it told me that they can't which (while still not being reasoned by the LLM) seems to support the spirit of your claim since it gave a correct answer while you (a presumed human that can reason) got it wrong.
Computers and software can be said to "understand", "think", and "reason" in their own way and informally people have always used those words in that context. Recently, software which has been trained on human-reasoned output is producing text that mimics reasoning well enough that it can be confused for the real thing, but nobody has been able to show that any reasoning (as a human reasons) is what's occurring.
If the output it produces is as useful to me as the output produced by a human with the magical and expensive capability to 'reason', why should I care?
What does this mean?
Maybe it'd be easier to try another definition:
2 a(1) : the power of comprehending, inferring, or thinking especially in orderly rational ways : intelligence
The same source defined intelligence as:
a(1) : the ability to learn or understand or to deal with new or trying situations : reason also : the skilled use of reason
And here we get the core of the issue. AI doesn't "think". It doesn't comprehend or understand what it does. There is no actual "I" in AI that didn't come from the people whose works were used to train it. At least not yet. I question if LLMs will ever be capable of anything more than producing a convincing affectation of the process used to produce the material it was trained on. I suspect that AGI will have to come from elsewhere. That doesn't mean that what passes for AI these days can't be useful, but I don't think it's capable of reason and as far as I know, nobody has proved otherwise.
Can a calculator comprehend arithmetic? Can it take a hold of a number (in a register, for example), and a second number, and add them together to get a hold of the result?
What is computation, really? When we design machines to do arithmetic, do the machines actually do arithmetic, or do they just coincidentally come up with states that we humans can interpret as a correspondence with arithmetic?
More importantly, would a rose by any other name smell as sweet?
If you put a problem into text, and give it to an LLM, and an LLM applied a series of higher order pattern matching to it to produce more text, and you read the resulting text and interpret it as reasoning about the solution to a problem, has the LLM reasoned? Does the calculator calculate? Or does it really matter?
Iran is one of the few countries that allow you to pay eg kidney donors. Guess who doesn't have a waitlist for donor kidneys?
And a poor person with one kidney less but extra health checkups can probably get better health outcomes than a poor person with two kidneys but extra health checkups. (This assumes that the kidney sale comes with some mandatory regular health checkups. Just like kidney donations do.)
In Iran, there is no tracking of donors and most donors don't want to be associated with the donation long-term, as it's connected to their poverty and the shame of their poverty. In the limited surveys that have been done many also noted that they haven't been properly informed about the risks of a donation.
Donation of blood and other replenishables is a bit different of course, but allowing open sale opens some bad avenues in terms of people effectively being forced to (which could carry risks they wouldn't normally want to take).
----
[1] Well, not practically…
Your two statements contradict each other.
In any case, we only have a finite number of days allotted to us on earth.
> [...] you can't donate a second time [...]
You can't repeat your 30s either.
Note that this analysis was performed by Dynonight, a rather bright blogger whose articles appeared several times on the HN front page. The vast majority of humans (I include myself here) probably wouldn't be able to achieve a result of comparable quality, even if it doesn't look that hard in retrospect.
LLM Deep Research can already exceed the performance of not-so-bright humans, but it is a quite different matter to outperform smart people like Dynomight. (I guess "research experts" isn't quite the right term here. The mentioned journalist from The Economist apparently was unable to research the topic to a similar degree, even though research is a main part of his/her job.)
I'm gonna go out on a limb and say that most other countries don't allow people to sell their blood for money.
E.g. Hungary allows you to give blood every 56 days (and allows selling it), Italy requires you to wait 90 days (and does not).
In the searches I have done Google's "Deep Research" has been better at providing primary data (or very convincingly fabricating).
OpenAI's version seems to more likely to give the answer that everybody thinks is true. For the blood example I could see it finding many sources that repeat the 2% claim, and accept that because everyone seems to agree, then it must be right. That's a mixed blessing in that maybe most casual users might want the commonly accepted answer, but when I have used the deep research tools, it has almost always because I know the answer everybody gives for a particular question, but I suspect it might not be based in reality. This makes my reason for wanting an automated deep research tool coinciding with the weakest area of the tool itself.
It's also been a bit eye opening how often commonly repeated but poorly founded claims, seem to turn up the same names of individuals, (or organisations, or individuals pretending to be organisations) as you trace them back.
Random hypothesis:
• equipment needs for uranium enrichment for Manhattan project in 1940s
=> US cornering the market on centrifuges (in both a "we buy them all" sense and a "we won't let companies sell them to other state actors in quantity" sense) for decades
=> US biomedical manufacturing of anything requiring centrifuging as a step, quickly outstripping that of all other countries
=> eventual global logistical dependence on US-based suppliers for such products
https://www.osti.gov/opennet/manhattan-project-history/Proce...
Centrifuge uranium enrichment wasn't developed to an industrial scale until the 1960s, and it first happened outside the United States:
https://world-nuclear.org/information-library/nuclear-fuel-c...
There are similar arguments made for kidney donation. Paying people a set amount kidneys is exploiting the poor, but most those who need kidneys are not well off and there is an enormous shortage causing massive shortages, so some argue that we should pay since society would be net better off.
Miele and Bosch makes great centrifuges for their washing machines. That technology is probably closer to what you need for blood, than the uranium enrichment equipment would be.
Not to mention how i get a cookie and the semi-goverment organisation charges >600€ for a baggie to hospitals. Someone needs to pay for that CEO's third house and car collection!
I used to operate a database for a large North American alcohol retailer. We had a problem with our data that said that 30% of the ring from a particular store saw the name "A GIFT FOR YOU" on the CC stripe. After months of bitter accusations of my database being incorrect, we flew someone out to investigate.
We found that a block away from the store was a plasma "donation" facility. In order to skirt various laws, when you "donate", you are given a prepaid credit card. The name on the card is "A GIFT FOR YOU". Donators then took that card directly to the alcohol store.
All of the data was correct; 30% of the ring from one store was paid by credit cards from the plasma "donation" facility with "A GIFT FOR YOU" on the card. A large reputational battle then commenced as the retailer thought of themselves as a high-end "wine and craft beer" store, when ultimately it turns out Budweiser pays the bills.
*edited for clarity
That said, following your Wikipedia link, it seems that a different mechanic is at play:
> "No product may be favoured over another, which in practice means that the beers are not refrigerated, since otherwise all beer would have to be refrigerated which is too expensive."
https://www.omsystembolaget.se/english/systembolaget-explain...
Source: new year's resolution to pick up at least 1 piece of trash per dog walk.
I live in Baltimore and I see a lot of people drinking singles out of a paper bag on the corner or their stoop. I've seen it on the bus, too.
Might be more of a hood corner store thing than a rural guy in a pickup buying a beer for the drive home type thing.
A few nice adults told us that it would stay cold a lot longer in a paper bag, and that a plastic bag wouldn't help. Both statements proved true…but most other adults would assume it was alcohol. They'd be scowling at us, telling us off. Sometimes we'd show the bottle, but few even need to smell it to be sure it wasn't alcohol.
Even then, most of them never started acting like they were the ones who had been in the wrong. All because some people were known to use a paper bag around beer bottles.
The only places that sell individual beers are convenience stores
Some of them are driven too, sure, but lots are also in city areas
Even so, people who only have a few bucks drive to the store where they can buy the beer, and drive home and drink it
If these were common cases, cops would lay in wait and nab them for open container.
I mean, they do. One of the big issues is coverage of enforcement. Here in Texas you see disparities where arrests in big cities converge on public intoxication, where in smaller cities you run into a higher ratio of open container because of cops having time to sit and watch stores.
There's also the rare mulled beer which is drunk hot.
(Joking.. but not really.)
While it is not easy to drink an open container on a bus or train, drunks will consistently stop in a nearby convenience store, have a cold one before boarding, and leave it there for others to clean up. I suspect that a high percentage of convenience-store sales for singlet cans and bottles may be attributed to pedestrians and transit passengers.
Aluminium is so comparatively high value, that it's usually in someone's interest to collect it.
Used glass is just bulky and cheap per gram.
And you're definitely right about recovery value. I believe, in Arizona, that glass has no redemption value at all, and so cans are the only logical choice.
There used to be a guy around here, a real character, called himself the Can Man, and he weighed about 350 lbs, and just rode the bus all the day, accompanied by an enormous bag full of cans that he would collect from every site possible. And he had songs he would sing and a little patter about himself, and whereas a can-collecting dude would normally be sort of smelly and repulsive, he managed to add to the "local color" and be somewhat endearing as he rode the bus around doing his can-collecting thing. I miss that guy!
And I buy them cold because I want them right when I get home. Not after an hour of cooling down in the fridge.
Edit: I just did a bit of research and didn't come up with a lot. I found this example [1] where the CEO's total compensation is $414k. That doesn't seem all that high...
[1]:https://givefreely.com/charity-directory/nonprofit/ein-57066...
In 2011 they made 260k€, which was at then time about 50% more then our prime minister made. Which is used as a norm to limit how much you are allowed to make when working for public services.
I mean to really get to a mansion and multiple holiday houses and a bunch of fancy cars you need to work the job for like 15 years, but still. In NL (where I think GP is from) earning that kind of money qualifies you as "seriously rich". Even more so if the salary got inflation-adjusted since 2011.
Someone making 3x a normal salary is not excessive by any means.
Entry level engineering (any type): 80k minimum
Entry level software engineering: 120k minimum
Multiply by 1.5x for high cost of living area (Bay Area, New York, Boston, LA)
Multiply by 2x for FAANG in HCOL area
Levels.fyi gives good estimates of software eng pay bands above entry level. Non-software almost always makes less :/
Either way, that’s kindof disconnected for the CEO of this bloodbank. Their opportunity cost is quite high. They choose to work for comparatively low pay, but have a positive impact on the world.
i do think it's more likely that the executive class has a vested interest in keeping their own pay high (not just CEOs, but board members as you've mentioned), and they've got a lot more class solidarity than the rest of us, though they'd never put it that way - it's just The Way Of Things, you know? divide and conquer.
i should maybe mention that i'm a public employee and my "opportunity cost" is also relatively high, but... hell is hot, and i want to be able to sleep at night.
If that correlation doesn’t exist, then investors should be able to easily drive CEO comp way down.
That's what we have eg corporate raiders for.
Since CEO is with them all the time, it follows they get the most.
There is a good theory on why CEO's are paid more than they're worth called tournament theory.
A person in an upper layer of a hierarchical organization is paid for the value they provide PLUS an amount that incentivized anybody below them to reach that layer. If everybody is paid just for the value they provide, they wouldn't have a reason to try and reach higher layers and they would minimize work. After all, we cannot assume the value a person provides goes up in higher layers.
This additional pay bubbles up in larger organizations to exorbitant levels for CEO's.
Their CEO makes $3 million, most of their executives above $400k.
https://projects.propublica.org/nonprofits/organizations/131... (look under Compensation)
Fuck this shit.
They also award them bonuses.
They are so bloated that they invest in PE funds with their spare cash.
Maybe not that high by SC/HN standards.
Presumably a European county given you quoted euros?
Looks like the EU encourages member states to encourage unpaid donations:
> Furthermore, Member States should take measures to promote Community self-sufficiency in human blood or blood components and to encourage voluntary unpaid donations of blood and blood components.
They say paying donors puts them & patients at risk, because it would stimulate donors' to lie about their health when donating and we can't be asked to test blood for everything. And the say it is unethical to pay for human tissues.
Except when they are selling it ;)
People have diseases they don’t know about and if people are ashamed by a disease, they’ll lie even when donating.
I just shuddered thinking that donations wouldn’t be tested.
Both of the key claims do seem to have evidence to support them:
1) Paying for blood shifts the socioeconomic distribution of the donor pool.
2) The socioeconomic shift results in more contaminated blood due to greater incidence of drug use and/or related blood-borne diseases that cannot all be tested for with high accuracy.
[0] https://www.goodrx.com/health-topic/finance/how-much-donatin...
edit: confused plasma donation with blood donation; it is still only every 2 weeks for plasma donations over here
Blood can be separated out into its plasma, red blood cells, and platelets by an apheresis machine. The machine cycles the unused components back into the donar so only one component is donated.
Blood plasma (~55% by volume), the amber colored water and disolved proteins, can be sold. Red blood cells (~44% by volume), and platelets (~1% by volume) can NOT be sold in the US by donars.
Most blood drives that you'd experience at school or in the workplace takes whole blood (so there is no need for the apheresis machine) which is more exhausting than if just one of the components was taken.
Source: an O+ blood donar with 50+ pints donated.
For anyone reading: Donar is "Old High German" for Thor.
A lot of plasma is also separated out of whole-blood donations and manufactured into all sorts of things. I don't know all of the end-user financial ramifications of this, but hospitals absolutely do pay (sometimes quite a pretty penny depending on rarity of antigens and antibodies) for RBCs and platelets (and plasma) from suppliers like the American Red Cross.
Purely anecdotally, I have heard stories of some donors being compensated extremely handsomely for their regular donations because of the rarity of their blood attributes - even being flown across the country and wined and dined to obtain their blood on top of thousands of dollars per donation.
https://podcasts.apple.com/us/podcast/part-one-how-tainted-h...
Note that a main point of the article is that it is not in the US' top 10 exports
Presumably that's a one time compensation, and you are talking about the annual recurring education budget?
That silicon valley skit, bryan johnson, and Theranos are all starting to come together now
Also interesting: "In 2023, total US goods exports were $2,045 billion, almost exactly ⅔ of all exports, including services."
That distorts supply/demand between countries
Sweat from China
Tears from everywhere else
Its just that the US has the largest population of these five countries. And generally the rest of the countries outside the five don't get sufficient plasma donations to make the drug products needed for their patients and have to import it.
The plasma can be separated out to different products to treat various diseases many of them genetic.
https://www.fda.gov/vaccines-blood-biologics/approved-blood-...
The only ones to hold accountable are the WHO and those countries that followed their advice. If more countries paid for plasma donations there would be more supply of these drug products available.
Interesting. Now I know why in a small town I regularly visit there's a high-street plasma "clinic".
Sad news.
As long as the demand is not steep. Since you can do it only a dozen or so times each year according to recommendations. (American Red Cross recommends every 28 days; private donors are not beholden to that)
Use either as a bargening chip and you are likely to lose them.
Point being: if it no longer makes sense to import 38% of blood from the US or whatever, it'll be imported from elsewhere or made locally, and that's pretty much the whole story. This is true of most goods and services, though not all.
Chip production is the very obvious and real hole in this argument. It'll take half to a full decade to get new chip fabs up to what TSMC's Taiwan fabs are doing now.
Quality steel production and machining is going to take time and investment as well.
I wonder when we are going to start seeing proper effects of all these tariffs in the market.
Disclaimer: I am the founder of DataLinks which in turn powers the searchtariff website
"Please show lots of digits"
Now, product groups for which data is most frequently and easily available is the 4-digit level, which is quite broad. If you look at the code 3002 in the HS classification system (of which there are many versions but we'll ignore that for now), you'll find a category, succinctly named:
> "Human blood; animal blood prepared for therapeutic, prophylactic or diagnostic uses; antisera, other blood fractions and immunological products, whether or not modified or obtained by means of biotechnological processes; vaccines, toxins, cultures of micro-organisms (excluding yeasts) and similar products; cell cultures, whether or not modified:"
https://hts.usitc.gov/search?query=3002
People new to trade data, especially programmers, with some hubris, tend to think this is way too long a category name to fit in a title or dropbox, so they chop it at the semicolon and call it good, resulting in "Human Blood" or similar. Better data sources tend to shorten these based on the real world percentage of the subcategories, e.g. see here "Serums and vaccines":
https://atlas.hks.harvard.edu/explore/treemap?exporter=count...
If you search for 3002 (Serums and Vaccines) in the US's exports in 2023 you'll see the figure 1.58%:
https://atlas.hks.harvard.edu/explore/treemap?exporter=count...
Which seems to me to be how they arrived at that incorrect number - some other website showing comtrade / us trade data with bad category names.
Lesson here: classification systems are hard.
https://www.iheart.com/podcast/105-behind-the-bastards-29236...
https://www.iheart.com/podcast/105-behind-the-bastards-29236...
I used to donate ever opportunity I could. I only stopped for medical reasons. I know there are a lot of companies that sell the blood, but it's still needed and can save a life.
[0] https://en.wikipedia.org/wiki/Never_Let_Me_Go_(2010_film)
Today I'll unsubscribe, I won't miss your preference for normie filler content..