A honeycrisp apple and a granny smith apple and a washington red apple are all apples.
You can make a number of comparisons among them. The criteria you focus your comparison on determines the outcome of the comparison.

Take away the human word "apple" and judge on their characteristics.

If you compare them by size, shape, kind of plant they grow on, which portions are edible/inedible, and the color of the outside skin, then your comparison would conclude that they are all the same object other than the outside color. Therefore, when deciding which apple is the best to buy for your kid's lunches, you would logically conclude "Oh my kid's favorite color is red, so the red one is the best one".

If you compare them by sweetness, firmness, aroma, price, culinary use, and color of the inside flesh, then your comparison would conclude that they are almost completely different objects and only have the inside color in common. Therefore, when deciding which apple is the best to buy for your kid's lunches, you could logically conclude "Oh my kid loves sour candies and prefers crunchy things over mushy things, so I won't get the red one, and the granny smith is cheaper than honeycrisp, so granny smith is the best one".

The map is not the territory. The map is just one way of describing certain characteristics of the territory for a certain purpose. Comparisons using the phrase "first world" or "apple" implies that the existence of those categories is enough to render all comparisons equal. But they are not equal. The result of any comparison is always determined by which criteria you choose to look at, and also which criteria you choose not to look at.

Yes, that's the point - it is a bad analogy because there are differences.
Hand-waving those differences away by using a vague fuzzy term like "first world" to presume that all other things are therefore equal, and you can take spending-per-capita on its own as a dispositive point of comparison, is a bad analogue, like hand-waving away the differences between alcohol spending in the USA vs theocracies.

1) What amount of operating costs should be covered? What number do we cap costs at, and who decides that number?
1a) How much cancer, surgical, EOL, etc. treatment is each participant entitled to? Does the limit exist? If so, where do you cut people off and why are those people's needs okay to deny? If not, how do you manage costs in an area where demand for goods/services is infinite, and the more successful it is at providing goods/services today, the MORE demand there will be tomorrow?
2) What is "a good wage" for each of the roles in a modern hospital? What number do we cap your wages at for each position, and who decides that number?
3) How large should the buffer be? What are "hard times" and how do we know when to spend the buffer?
4) What are the metrics by which we know we have reached our target of "reliable and resilient"?

I do not disagree with that diagnosis in a significant way, other than to quibble over semantics of the phrase "the majority of its beneficiaries". The majority of its beneficiaries are me and 300 million other US citizens - both on the consumer side and the provider side. But I could see someone making the case that the largest chunk of the dollar-value benefits go to the corporations who run the system. Even then, it's not a clear boundary because the entire global economy has become parasitically/symbiotically dependent on the speculative stock/investment industry. If we hand out corporate death penalties to all the biggest worst businesses today, we will all be starving in the streets by the end of the month; forget the modern invention called retirement and prepare yourself to labor all the way to the grave like almost every human who lived before 1930.

Consumption, whether per-capita or total, is not sufficient to compare different societies' use of goods.

Spending on alcohol per capita is orders of magnitude higher in the USA than in, say, Iran. Is that because people in the USA are "overpaying" for alcohol? Or do we need to consider lots of other factors to make a valid comparison?

Similarly, I'm sure healthcare spending per capita is significantly higher in the USA than in, say North Korea. Does that comparison mean the citizens of North Korea have better health and wellness than citizens of the USA? Or are there a variety of cultural, regional, economic, regulatory, technological, etc. factors that both determine that amount of spending and also determine whether people feel happy with the care they receive?

It works, of course, because the stories with the highest comment count the past 3 years are consistently: climate change, energy generation/consumption (fossil v nuke v wind/geo), AI-yi-yi, and cryptocurrency.

You never have purchased cosmetics, skin care, vitamins and nutritional supplements, diabetic supplies, baby supplies, pet supplies, etc.

It is trivial to get to $800 in a Walgreens, with fewer than 20 items, without needing to scour the store for obscure items. Just throw a few bottles of CoQ-10 and Glucosamine in your handbasket, then stop by the lotion aisle to get a set of whatever the newest ANooYoo[tm] Rejoovenating RetinaCollagenauric Peptide Facial Anti-Aging regimen is, and you can hit $800 in about 5 minutes.

Yes. But, paradoxically, only for the very very weak-minded and the very very strong-minded.

Most modern people don't think about it that closely, and noncommittally hover at the upper end of the first condition.
A significant minority fully outgrow it and then plateau, content to splash around in the mental pleasures that come from opposition to their predecessors.
Only a very tiny minority continue their intellectual growth long enough to achieve the second condition.

The human intellect is heavily flawed in discerning truth in situations where it's easy to be right for the wrong reasons and wrong for the right reasons. Religionists and atheists are brothers in the quintessential human tendency to respond to paradoxes and the inherent unsolvable incompleteness of knowledge, with leaps of faith that give them the mental relief of feeling vindicated in their absolutism. Which is preferable to admitting that our intellects and perceptory structures are mere evolutionary kludges which are as capable of encompassing the universe's reality and making positivist judgments about it, as the intellect and perceptory structures of a honeybee are capable of encompassing and making positivist judgments about the meaning of an airport they fly through looking for pollen, or a live performance of "Hamlet" in a theater whose walls they're nesting in.

But at, what, ~30 years old in the storyline, the only reason he even has muscle fiber left to be acupunctured is that The Matrix creche must also have been burning energy inputs to do the same thing for 30 years. A newborn baby put into a coma, who then never physically moves for the next 30 yeas does not automatically develop the muscles to be a walking talking adult human being, just because time passes. The body's physical maturation is stimulated by continuous, escalating activity and nutrition. What reason would the Machine Intelligences have to expend energy providing all the extra inputs necessary to keep the body potentially red-pillable, when they could cut us down to brain-in-a-jar battery packs at a much higher efficiency rate.

If you're going to include all prior inputs, then we need to calculate the cost of 300 years of Industrial revolution which led to the creation of computers and software.
The technology for creating humans is an already-sunk cost from millions of years ago that has fully depreciated out.

Also, every LLM is the offspring of the all the human beings who wrote its code, installed its hardware, manage/maintain the processor-farm and network connections which allow it to process tokens, manage/maintain the businesses who pay for its infrastructure, etc.. Given that the modern technological civilization only works because a few billion meatsack human beings labor daily to make it work, your widening the scope of the comparison doesn't end as neatly and clearly as you state.

An enhanced LLM that churned through tokens for five hours, versus a human brain that works on the same problems.

Anyone here have idea how the energy consumption of this LLM processor-farm compares to the energy consumption of the next-place human contestant during the same time period? If Andy Weir re-wrote "The Martian" with an LLM-powered drone as the main character, how would the calculation of potatoes-versus-poop-versus-water change?

To me, the obvious plot hole in "The Matrix" series was the absurdity of the notion that bio-batteries - human brains and bodies - could magically violate physics and provide more energy/processing power output to The Matrix than the inputs that would be required to keep the nutrient/stasis apparatus running. Billions of people kept alive in a coma, apparently without significant muscular atrophy or any damage to brain development, because when people take the red pill and escape The Matrix they are walking and talking a short while later. The infrastructure required to feed, chemically stimulate, neurologically stimulate, and maintain homeostasis for billions of meatsacks would be prohibitively more expensive than just burning those energy inputs to directly power the Matrix.

Yes, if we put a billion monkeys on a billion typewriters (the training set and subsequent LLM token-producing functions) for a billion years (processor cycles in an AI farm), then they can produce Ye Compleat Works Of Shakespeare. Or, well... just give one monkey some porridge and water for a few decades and he will also produce Ye Compleat Works Of Shakespeare because he's, like, you know... Shakespeare.

I don't blame you. It's a lot. To be honest, I got under-stim boredom halfway through listening to my own inner monologue as I typed it. And I couldn't be bothered to do significant editorial review because I wanted to hit that Submit button as fast as I could to get the hit and speed my way to the next hit.

Try simultaneously binging a shiny-new-and-then-cancelled Hulumazonflix series in order to make my comment neurochemically tolerable (barely). If that doesn't work, binge harder or scroll harder.

Engagement farming doesn't merely sell advertisement.
It habituates cognition to a higher base level of engagement-stimulus, and THAT part works regardless of whether you eventually click through or buy a product. Simply seeing a clickbait headline stimulates the brain even if your higher-level critical thinking recognizes the bait and reminds you not to click the link. Your reaction of eyeroll, disdain, disgust, annoyance as you scroll past it is still a state of neuro-endocrine excitement. And every public and private experience is now being reshaped around engagement.

You simply cannot sustain that throughout entire narratives in long-form novels. If everything is the most important thing, then no thing is important. Narratives need dynamic range. The longer the narrative, the larger the range.

The movie "Run, Lola, Run" was a deliberate exercise in maintaining the constant tension of narrative excitement. But even that film only worked because it was an outlier within a storytelling medium with a wide range of immediacy levels. if every film was at that same intensity, the collective audience would become less interested in film as a medium, even if they couldn't explicitly explain why. It's why narrative arcs like the MCU can't perpetually dominate the field. You can only escalate the "existential threat to the country planet galaxy universe" so far before you escalate yourself into a corner.

You either have to balance the excitement with tedious characterization backstory or throwaway "monster of the week" episodes to preserve the value of the single long-form payoff, or you have to abandon the single long format buildup and shift toward a series of constant lower-level hits which must necessarily be kept short and narratively isolated from each other to preserve their punch.

In this unacknowledged global epidemic of tech-caused dopamine tolerance, we have chosen the latter.
The activity hasn't changed, but the functional payoff of the activity has. It's not that people are reading less, it's that what they are reading is less durable. It has to be, in order to maintain the stim level of each short snippet.

The article says "reading for fun is plummeting". Well, when people scroll 8,000 words on their socials, are they not reading for fun? I'll answer my own question -- no, we are not, because our cognition is being reprogrammed on a massive scale. The nature of "fun" has changed from more of a satisfaction-completion model to a stimulation-maintenance model. Ask yourself whether you ever feel satisfied or fulfilled at the end of the night after interstitially side-scrolling your feeds for two hours. Have you finished it? Do you ever get to any sort of end of the feed and feel that "Ahhhh.... now I see how it all came together" cognitive payoff you used to get from finishing a novel? Even when you stop scrolling, is it because you have reached fulfillment and enrichment, or is it simply because the time has come to force yourself to darken the screen and go to sleep just so you can make it through another day of work?

Stories like this one always result in people upping initiatives to push books on kids, as if access to books is still as rare and challenging and elitist as it was in 1897. In fact, the supply/access to books in 2025 is so pervasive that the monetary value of individual books is approaching zero. Which is why we now have hundreds of thousands of people with a "Little Free Library" in their front yard or church lawn or local park where they literally give millions of books away for free to anyone who wants one.

But access is not the cause of this story. If "reading for fun" is plummeting, it isn't because people are having a hard time finding books, it's because people are having a hard time reading books, because it simply isn't fun anymore. The cognitive nature of "fun" has changed, so when our brains are looking around the local environment for sources of "fun", the "fun" provided by long-form reading is being compared to this new level of "fun" which is immediate and infinite and probably already in your hand/pocket right now.

As Alanis Morrissette sang: "I've got one hand on my cellphone, and the other one is scrolling my cellphone screen."

It's dopamine all the way down.

The deep tragedy here is -- no one can sue the person who trained the bot to become a suicide coach, because that person is no longer alive.

"AI" in its current form is a mirror. The more you interact with it, the more it becomes you. You use your words to incrementally, continuously coach it to display on your screen more words which could be your words and thoughts. This person took a mirror and stared into his Self-abyss so attentively that his abyss began staring right back into him, and talking to him.

That hurts, it's awful and fragile, and I sympathize with the parents who are trying to make the world provide them with closure for a wound which - in my experience - is fundamentally impervious to closure. Over time your grief dulls or the questioning and guilt becomes a manageable part of you. It's like a corneal defect, or that splattered bug on the edge of your windscreen -- never completely invisible, hovering just out of conscious thought in your emotional periphery, but can fill your field of vision quickly if something causes you to turn and look at it again.

The suicide of a loved one is a violently chaotic shredding of the fabric of your world.
There is no closure. No resolution.
Any hope of resolution died when they did.
You will carry that jagged scar for the rest of your life.