Paul_Clayton

"higher prices for the consumer" can include lower value at the same price. Size reduction seems a common method for certain commodities. This may result in reduced consumption (e.g., a consumer buying one package of ice cream every other week), as well as increase customer dissatisfaction when the change is noticed, and it can increase packaging cost per unit weight/volume.

Other ways of reducing value are possible such as reducing quality control effort, reducing quality of inputs, and reducing manufacturing costs in ways that are known to reduce product quality.

Pushing costs to effectively underregulated externalities can also avoid price increases.

It is also sometimes possible to increase efficiency. Even a long term commodity can have potential for efficiency improvements that were considered not worth exploring under stable pricing pressures. (I suspect value reduction is easier and much faster than efficiency improvement.)

Sadly, reducing value can have a disproportionate cost to consumers. Reducing manufacturing costs for a Watchman's boots by 20% may reduce the lifetime of such by 30% and reduce the quality of use by 50% (which may be related to Samuel Vimes' theory: https://en.m.wikipedia.org/wiki/Boots_theory ).

Innate talent can also be tied to one's sense of identity, which makes failure more overwhelming. (If one does not _really_ try, failure does not feel as crushing.)

Just an expectation that something will be easy (without a strong tie to identity/sel-worth) can make failure more painful.

Easy successes can also lead to not developing progress-enabling skills when in a "friendly" environment (e.g., an academically gifted person not learning study skills and disclipline before college). When the innate skill and casual training is no longer enough to meet expectations, there is not the emotional reserve and external support to develop the meta-skills.

Failure aversion and lack of self-discipline is somewhat independent of "work ethic"; a person terrified of failure can work very hard at easy tasks or tasks with results that lack internal or peceived external judgment in part because such feels so much better than not really trying.

Sadly, a "safe" activity can be "ruined" by a person's well-meaning compliment, that introduces expectations to the activity. (Weirdly, indirect compliments seem significantly less problematic; "these decorations look really nice" can feel acceptable even when the person knows one did them while "you did a really good job on the decorations" can feel crushing by setting a new higher baseline of expectations and/or introducing self-doubt because the person is just being nice.)

This reminds me of a passage in Lois McMaster Bujold's A Civil Campaign (Cordelia Naismith Vorkosigan is speaking to Mark Vorkosigan):

Her smile tilted. "Mark, you don't pay back your parents. You can't. The debt you owe them gets collected by your children, who hand it down in turn. It's a sort of entailment. Or if you don't have children of the body, it's left as a debt to your common humanity. Or to your God, if you possess or are possessed by one."

"I'm not sure that seems fair."

"The family economy evades calculation in the gross planetary product. It's the only deal I know where, when you give more than you get, you aren't bankrupted—but rather, vastly enriched."

I received the impression that being closed on Sundays was part of corporate culture, a statement of valuing something other than maximizing value extraction. A poster here commented that the franchise agreements were unique in the poster's experience in having a flavor of partnership. Customer service seems to be emphasized.

B&H Photo seems to have a similar (even greater?) customer service drive. More than one poster here has commented that being closed on Jewish holy days (including Shabbat) is not a major deterrent given the quality of customer service and products.

Keeping one's word even when it hurts is a mark of virtue and integrity means such virtue is expressed in a breadth of life activity. Miracle on 34th Street hints on the advantage of "having a heart", even if Mr. Macy saw it as a corporate gimmick.

A McDonalds being closed one day a week because the managers like outdoor activities that day (or to try to achieve a Chick-fil-A effect) would see the ethos behind that behavior informing other behaviors.

Marketing can support a false reputation, but trust is important to social function. Betrayed (or mistaken if you are Roj Blake) trust wounds society.

The Catholic Church had similar tradeoffs with Latin, though I suspect the language and style were less motivated by majesty (though bias of use by the educated might have entered early — I am ignorant of the history). The New Testament Koine (Common) Greek was similarly a lingua franca. When the once common language is no longer broadly used, the language can become a class-oriented separating factor.

Even more recent translations seem to retain significant similarity in a lot of "famous" texts (e.g., the Beatitudes — people also seem to use the archaic pronunciation of "blessed" as two syllables), presumably to ease acceptance of the change. This hints that some commonality is preserved. (Some words are also jargon, so not modernizing the word is more reasonable.)

Story outlines and concepts can also be preserved even though the "poetry" of earlier versions is lost in translation. Yet as contexts change even concepts may be less understandable and shared; "go to the ant thou sluggard" may be unclear not merely from language but from unfamiliarity with concepts. Aesop's "The Ant and the Grasshopper" has lasted thousands of years, but is not a fundamentally human metaphor and even the human concepts diligence and foresight can have different cultural tones.

"A sluggard is someone who does not work hard." "Oh, you mean someone who works smarter not harder?" "No. It means someone who does not accomplish much." "Oh, you mean someone who is burnt out?" "No. It means someone who chooses not to do things that are profitable." "Oh, you mean someone who has recognized the futility of striving for accomplishments and has learned to be content with a simple life?" "No!"

For some reason — I am musically incompetent — "The Beach" really felt like the soundtrack was very emotive in a way that fit the story.

The stately feel in "Sleepytime" may have fit the solar system motif and provided contrast to the earthy elements of the story.

I liked "Sleepytime" in part from the expression of letting go while being faithful. Bingo's letting her rabbit go did not mean a complete separation but an expansion (as a few additional rabbits came to help later) similar to a parent allowing a child to grow up and leave while affirming that the child is loved.

Bluey also has some similarity to Calvin and Hobbes in that some fantastic elements are not explained. E.g., in "Sleepytime", do Bluey and Bingo have a psychic connection such that they shared part of the dream or is there imagination/unreliable narration or some third option?

Listing all of the creators from whose attribution-licensed works an LLM (potentially) derived an output would seem to satisfy the letter of such licenses, but it is not clear that such would satisfy the spirit (which seems to assume a stronger causal link and a more limited number of attributions). If creators can be grouped outside of the creator naming explicitly associated with the works, this could degrade into "this work is derived from the works of humanity"; however, listing all human beings individually does not seem _meaningfully_ different and seems to satisfy the attribution requirement of such licenses.

From what little I understand of LLMs, the weight network developed by training on a large collection of inputs is similar to human knowledge in that some things will be clearly derived (at least in part) from a limited number of inputs but others might have no clear contributor density. If I wrote a human "superiority" science fiction story, I could be fairly confident that Timothy Zahn and Gordon R. Dickson "contributed"; however, this contribution would not be considered enough to violate copyright and require licensing. Some LLM outputs clearly violate copyright (e.g., near verbatim quotation of significant length), but other outputs seem to be more broadly derived.

If the law treats LLMs like humans ("fairly"), then broad derivation would not seem to violate copyright. This seems to lead toward "AI rights". I cannot imagine how concepts of just compensation and healthy/safe working conditions would apply to an AI. Can a corporation own a computer system than embodies an AI or is that slavery?

If the law makes special exceptions for LLMs, e.g., adjusting copyright such that fair use and natural learning only apply to human persons, then licensing would be required for training. However, attribution licenses seem to have the above-mentioned loophole. (That this loophole is not exploited may be laziness or concern about admitting that following the license is required — which makes less openly licensed/unlicensed works poisonous.)

If the purpose of copyright is to "promote the useful arts", then the law should reflect that purpose. Demotivating human creators and the sharing of their works seems destructive to that purpose, but LLMs are also enabling some creativity. Law should also incorporate general concepts such as equality under the law. LLMs also seem to have the potential for power concentration, which is also a concern for just laws.

Perhaps I am insufficiently educated on the tradeoffs to see an obvious solution, but this seems to me like a difficult problem.

The enthalpy of vaporization of water (at standard pressure) is listed by Wikipedia[1] as 2.257 kJ/g, so boiling 462 grams would require an additional 1.04 MJ, adding 26 seconds. Cerebras claims a "peak sustained system power of 23kW" for the CS-3 16 Rack Unit system[2], so clearly the power density is lower than for an H100.

[1] https://en.wikipedia.org/wiki/Enthalpy_of_vaporization#Other... [2] https://cerebras.ai/product-system/

Colonization would, I think, not require economic justification. Using transportation and communication barriers to facilitate isolation or at least localization may be important to some people.

With vast discretionary spending, people would be able to sacrifice some pleasures for pleasures related to achieving a long-term impractical goal. Space travel does not have to be the most efficient means of enjoying life to be a human choice.

A colony in Antarctica would be much more practical than one on the Moon, but a Lunar colony is more romantic.

If material trade is not worthwhile, conquest may be somewhat discouraged. (Some people might wish to rule all of humanity or convert/exterminate all heretics, so no distance would be perfectly safe, but a perceived threat in Antarctica would be easier to eliminate than a perceived threat on the Moon or in the Large Magellanic Cloud.)

Another problematic aspect is if one has poor eyesight and wishes to use a larger font size. One ends up having to scroll horizontally for each line of text for single column pages. For two-column pages, one has to scroll back a page after reading the first column.

Sometimes one can use a landscape-oriented display to avoid horizontal scrolling, but even if the same word count fits on the screen I seem to be annoyed by the low line count.

Providing large type and huge type PDFs would not entirely solve this problem as sometimes even one with poor eyesight might prefer a smaller font for scanning/skimming. Having to acquire two PDFs and switch between them based on mode of use seems suboptimal.

Fixed paged presentation has significant advantages for familiar reference material; some people seem to have a spatial memory that makes finding specific content by flipping pages faster than trying several search phrases (with the occasional benefit of serendipity).

Poetry often benefits from not reflowing lines and page breaking within a stanza is often more jarring than within a paragraph of prose. Yet a reader might prefer inferior typography over having to use a magnifying glass or carry a very large display.

One might be able to get some of the advantages of paged media for figures and tables by having header (or footer) pop-up links to such content when it is on the same "page" as the displayed text. This is not as low effort as moving one's eyes, but it might be better than inlined presentation on a small (relative to font size) display.

Even with print, there would be times when breaking the text to fit a figure is more disruptive than having the figure on a separate page. There would also be times when all the relevant figures would not fit on the same page as the related text.

Having a separate booklet of illustrations might make going back and forth between text and illustration easier, similar to having a lexicon or commentary open while reading. However, that also introduces position tracking in another book and other inconveniences.

Even when my vision was better, reading academic papers distributed as PDFs (usually 2-column) on a computer screen was less enjoyable than reading similar material in a reflowable format. Academic papers also do not seem to benefit as much from pagination as other writings.

Die yield can depend on redundancy and sensitivity to process variation.

Column/row redundancy for SRAM arrays is common and multicore designs provide core redundancy. Theoretically even out-of-order scheduler entries and functional units could be disabled. SIMD width variability could provide another means of tolerating defects and/or variation while providing a sellable result.

If a usable/sellable die only needs to reach half of the best frequency or half the best case energy efficiency, yield can be higher than if nothing less than 95% of best is worthwhile. If better dies can be sold at higher profit, the economics change. Durability is also a variable that can be tuned (e.g., "The BubbleWrap many-core: Popping cores for sequential acceleration" https://scholar.google.com/scholar?cluster=13412927692517066... ).

Yield does not seem to be simply a matter of defects per square centimeter. At least so it appears to this computer architecture enthusiast.