That is abstraction of the implementation of the tool, not the output.

Producing outputs you don’t understand is novel

> A recent study shows that on any given day, just 12% of people in the US account for half of all beef consumed in the US.

This phrasing strongly suggests it’s not the same 12% every day. In which case… it’s probably not that noteworthy.

The phrasing strongly suggests exactly the opposite. Essentially, the whole framing of the linked guardian article is that there is a specific population which are the "disproportionate beef eaters".

And from the study linked, that framing/suggestion would be incorrect (at least for the numbers given). "the 12% are not the same every day" is an accurate interpretation. They asked about what people ate _yesterday_...

Again, the whole premise of the article is that there really is such a thing as disproportionate beef eaters (DBE), and it spends time talking about this group explicitly. So the wording doesn't suggest otherwise, it explicitly suggests this is a real group.

Regarding the study this is a both can be true situation. There can be (1) a population who is disproportionate in their beef eating, and (2) a study about 12% doing the most on any given day can count in favor of that group being real and (3) not everyone from the daily 12% is part of the DBE group. It's more likely a venn diagram overlap, and where it doesn't overlap, people who aren't part of the DBE are incidentally in that 12% while being closer to average in the aggregate over the longer term. Those facts can all sit together comfortably without amounting to a contradiction.

They can both be true but 12% of people eat 50% of beef on any given day implies that way more than 12% of people eat 50% of beef on any given year.

Like, it’s probably something closer to 40% than 12%.

The phrasing you’re looking for is that 12% of Americans consume an average of 50% of beef consumed every day.

By saying “on any given day” you are suggesting it’s a different 12%. The article does confuse this by identifying cohorts that eat more beef. But it’s a tautological label based on the survey data. They identify some correlates, like being a 50 something male. But there are males who are 50 something that don’t eat any beef. They’re not included in the 12%.

The 12% is just the outcome of the sample. It doesn’t mean they’re a consistent cohort.

Example:

* on any given day x million women give birth

* there are x million women who give birth every day

is it normal, in the USA, for half of all people to only eat beef once every 8 days?

They also found a demographic correlation, which isn't easily explained by random sampling.

> is it normal, in the USA, for half of all people to only eat beef once every 8 days?

Thats not the implication of 12% of Americans eating 50% of beef by consumed by all Americans that day.

If I had to make up some numbers it’s probably that, on any random day, 12% of Americans ate 50% of the beef (a large burger), 28% of American ate the rest of the beef (bit of lunch meat), and 60% of Americans did not eat any beef.

I feel like this article is littered with suspicious statements

Like this one:

> In fact, institutional homebuyers (those who bought 100+ homes in a 12-month period) didn’t even reach 2.5% market share at the peak level in this data line, which goes back to the start of the century.

I don’t know how to evaluate this. I doubt this analysis rolls up subsidiaries. So what does it really mean for an entity to own 100+ units? Is that actually something we care about?

Imo only thing people need to give a shit about is whether a house is being bought to be lived in.

Platform economics create monopsony problems. If you don’t play ball with steam, you don’t get access to most customers. End of story. These things are winner take all.

Imo 30% is disgusting and needs regulation.

Edit: see also, credit cards

If the US tech companies stop behaving this way, maybe.

How is the EU tech company lack of growth related to fining companies for not obeying the law?

Yes, Europe is a laggard in tech, but I don't see any relationship here. Even if they wouldn't fine these companies, EU would still lag, and now that they are fining them, EU companies are not at an advantage, nor growing faster.

Europe just doesn't have the "move fast and break things" mentality because we don't want things like privacy broken. At least not without the user's unpressured choice which is what GDPR is all about.

If we allowed the same kind of unrestricted development we'd have more money and growth but we'd be just like the US. Which I personally don't want for sure. I'm glad to be living here. It's not all about money and economy.

The US is in the middle of a recession if you exclude the AI bubble. Even if you include the AI bubble it's barely avoiding stagflation. I'm not sure "growth and innovation" accurately serves as a contrast between the US and EU tech companies right now.

Or the US tech companies could abandon EU markets

You must believe that US companies are trying to enter and stay in hostile markets out of the sheer kindness of their hearts. Have you considered that not being present in the second biggest market by GDP may actually be a massive liability by creating a massive opportunity for competitors that will be far better adapted to stricter regulatory conditions? You could just as well advise US car manufacturers to stick to building cars like the Cybertruck and ignore markets that consider it unsafe.

They could, it could be a blessing for competitors in the EU.

But they won't because the EU is a huge market and money speaks, while that happens they need to comply with the laws. Stop breaking the laws and you stop being fined, it's pretty simple for multi-billion/low-trillion market cap companies, innit?

It would be terrific if this happens. Can give room for alternatives to grow in the EU. Even the rest of the world would love it.

So far only China has managed alternatives - and only thanks to govt exclusion. US behemoths just eat everyone else up - even in the global South.

I’d love alternatives that work well, but having used the said Chinese ones, I got no choice but to stick to the behemoths. Telegram may eat a bit into the messaging dominance, but that’s it.

I’m sorry to disappoint you but the EU is unable to create any usable alternatives to US tech chiefly due to lack of SWE talent (among other things). Anyone remotely competent sees the 40k senior SWE salaries offered by European tech companies and immediately crawls through glass just to work at a mid-tier company in the Northern California area of the United States.

Once they pay "modern" US health insurance (esp after a layoff) and also need to raise a family, the vast majority will crawl back through lava.

Still no - PPP (that’s after expenses such as food, healthcare, housing, etc) is significantly higher in the United States.

I believe that would be true (after food, housing, healthcare, taxes, child-care, etc) only for a very narrow band of senior SWE's. And you are still not considering employment protection. And for junior or mid-level SWE's, not at all true for the overwhelming majority.

This is true not only for SWEs, but for the median worker in the United States.

Huh no. I'd never work in the US. I won't even visit there as long as the current regime is in place (and the mandatory social media declaration, which I believe is more bipartisan).

I even moved to a lower wage country in Europe even to a pay cut, money isn't everything. Quality of life is. I won't live in a country that is anti-LGBT and with such a culture glorifying toxic masculinity. And at the same time giving a huge middle finger to the world by having the most polluting country in the world per capita quit climate change reduction efforts.

I don't think you understand how badly Trump has destroyed the reputation and goodwill of the US to the rest of the world in just one year. Everyone I know is actively trying to disconnect from US products and services (though admittedly I am in more activist circles)

And salaries here are a lot higher than that. Even here in a lower-wage country. Also, I don't need a car where I live which scraps a whole category of expenses, healthcare is free and I have protections in case I get fired.

I’m surprised that Iran can contemplate affording this. There must be such immense losses of all the land, homes, and capital assets in Tehran. And then operational costs of moving people around, building new homes, etc.

$100B is such a high number that it becomes funny money but… idk, doesn’t it still feel like a lowball in terms of losses?

Seems dubious to me that smoking explains any of these things. I would guess just a cultural cofounder.

Obesity and fentanyl would be my guess.

Smoking is a major health risk, so it makes sense to me that it's a major factor. Not sure why you'd want to remove it as a confounder.

I think parent comment was pointing to lack of establishing a causation link. The finding in their abstract is extrapolated by statistical inference. For example smokers tend to drink more etc. The paper does take such factors into account. Personally I wouldn't jump to such a strong conclusion from statistical inference because it closes the door on other factors that might be even stronger when combined together. The paper reflects motivated reasoning more than a discovery outcome. That said, smoking is of course a major health risk, I am just pointing at the research approach.

Smoking is a major health issue, but it's barely a driver of midlife mortality. Smoking tends to kill you later.

When my city banned public smoking, there was an immediate and statistically significant decrease in heart attacks and strokes.

But probably not for the reason you think.

What was the reason?

Smoking is bad for you, but it's unlikely to kill you in midlife.

In the paper they claim it matters for midlife mortality too:

> People who start smoking at age 18 begin to exhibit higher mortality several decades later, with particularly large effects beginning at ages 45–64 (Lawton et al. 2025). A health-capital model allows the mortality rates of older persons to be determined not only by their current smoking behavior but also by smoking in earlier years. In the United States, smoking rates started falling for college graduates earlier than they did for the non-college population.

...

> [...] with rapidly improving treatments and screening for lung cancer (Howladeret al. 2020), the major impact of smoking over the longer-term—particularly for people aged 55–64 arises from other more-common tobacco-related diseases such as chronic obstructive pulmonary disease (COPD); cardiovascular diseases such as strokes, aneurysms, and heart attacks; diabetes; and other types of cancers (Carter et al. 2015). Perhaps more surprising is that past county-level smoking rates are highly predictive of deaths of despair. This finding, however, is consistent with an emerging literature in biology that points to a causal influence of smoking on drug addiction [...]

It is associated with other adverse behaviors like a sedentary lifestyle. Smokers aren't generally athletically motivated.

Actually, later in their paper they say: "Although we have argued for a causal role for smoking in generating these patterns, the growing mortality gaps still seem too large and the causes of death too varied to blame the patterns on the adverse health effects of tobacco use alone. As noted above, smoking is likely to play a role in amplifying the impact of other factors adversely affecting midlife mortality, such as the marketing efforts by opioid manufacturers targeted to areas with high rates of smoking-related illness, coupled with epigenetic changes making smokers more susceptible to opioid use disorders. Still, the strength of our findings that smoking is predictive of spatial trends in midlife mortality points towards different mechanisms needed to explain the troubling trends that have unfolded since 1990."

> manufacturers targeted to areas with high rates of smoking-related illness

Oof

> epigenetic changes making smokers more susceptible to opioid use disorders

This one seems… a bit mystic to me. I would have been much quicker to suggest that a psychological propensity to start smoking mirrors a propensity to start using other drugs vs. arguing for emergent effects of cellular behavior.

Yeah it seems silly.

Why is college the primary group factor…? Is there some sort of health effect of sitting college classrooms for 5 years? Seems unlikely.

College education is highly associated and predicted by income/access to wealth.

Wealth inequality seems like a more likely explanation. Not seeing how they controlled for that across college vs non-college groups.

If you can only check pairings one at a time I’m not sure it’s possible to do better than greedily solving one person at a time.

So, for 10 pairs, 45 guesses (9 + 8 + 7 + 6 + 5 + 4 + 3 + 2 + 1) in the worst case, and roughly half that on average?

It's interesting how close 22.5 is to the 21.8 bits of entropy for 10!, and that has me wondering how often you would win if you followed this strategy with 18 truth booths followed by one match up (to maintain the same total number of queries).

Simulation suggests about 24% chance of winning with that strategy, with 100k samples. (I simplified each run to "shuffle [0..n), find index of 0".)

Agreed. There's an argument elsewhere about how a truth booth can possibly have an expected return of more than 1 bit of information, but in reality most of the time it's going to give you way less than that.

I saw an episode of this and felt the contestants didn’t seem that interested in winning the money. Just romance. I was curious how suboptimally they tended to play.

Everything is lined up for sub-optimal play.

For a start, the setting is an emotive one. It's not just a numeric game with arbitrary tokens, it's about "the perfect romantic partner." It would take an unusually self-isolating human to not identify who they feel their perfect match should be and bias towards that, subconsciously or consciously. We (nearly) all seek connection.

Then, it's reality TV. Contestants will be chosen for emotional volatility, and relentlessly manipulated to generate drama. No-one is going to watch a full season of a handful of math nerds take a few minutes to progress a worksheet towards a solution each week coupled with whatever they do to pass the time otherwise.

I'd watch a game show where you put a variety of math nerds on each team and watch them argue about the optimal strategy. Who's strategy will win? The quant analyst or the bioinformatician? Tune in next week!

Reminds me a bit of Devils Plan, or other similar reality game shows in Korea / Asia.

I saw a couple devils plan and felt they were trying to imply Mathematical scheming without substance. There was one guy who kept talking about “science” but then his “game theory” for a game of mafia was “social distancing” like Covid…

Watched a fantastic film about this on the plane a few years ago, "Liar Game - Reborn". There is some fairly sophisticated logic puzzling and scheming going on (see e.g. sample illustration https://imgur.com/a/0AOb67G from an interlude about 50min in where there are 3 groups of people who mutually distrust each other, each know a secret collection of 3-4 integers unique to their group, and want to deniably pass share integers with each other which are "not my team's". Another participant watches what happened and realizes in retrospect this is how the info was shared.)

A lot more upbeat than "Alice in Borderland".

We need a ManningCast version of the show. For those unaware, ManningCast is a show following an NFL game with special guests and nontraditional commentary and analysis. Think of it kind of like having the Mannings in the living room while watching an NFL game.

In my hypothetical version of "Are you the one?", the math nerds would be giving commentary and explaining the math behind how they'll solve "Are you the one?" while also hilariously explaining how foolish the contestants' theories are.

Need to find out their psycopath screening technique

Charlie Brooker did a good bit (in 2007!) about picking the right candidates for a mock reality TV show: https://youtube.com/watch?v=NGkJxju3uKo

Just Ask them to describe Shannon Entropy. If they start talking about information they are out, if they start talking about their crazy cousin they are in.

> No-one is going to watch a full season of a handful of math nerds take a few minutes to progress a worksheet towards a solution each week coupled with whatever they do to pass the time otherwise.

Um, what about those of us who watch Blood on the Clocktower streams?

Both of you are the exceptions that prove the rule?

That's because the real game is occurring both before and after the show in modern reality tv competitions. The goal is to be entertaining and get social media followers and potential invites to further reality tv shows.

I’m not really following. But if you’re told that one of A, B, or C is true; you learn more by being told A is True than if you learn D is True, no?

Yes, you learn more than 1 bit in that case. However, if you are told A is false, you still don't know whether B or C is true, so you gain less than 1 bit. Assuming A, B and C all have equal probability, your average/expected information gain is <1 bit.

If you ask the question "which of A, B, or C is true?" then you're not asking a yes/no question, and it's not surprising that you expect to gain more than 1 bit of information.

but that’s all consistent. “Expected” gain is less than 1 for the truth booths and sometimes > 1 for actuals; and is > 1 on expected value of the match ups, which aren’t binary questions.

Sure, and the issue is that the article says "Suppose we have calculated the expected information gained by potential truth booths like below:" and then lists some values >1

edit: just saw that the article fixed this recently, and the values are now <1