I recently swapped from the F-Droid version of AntennaPod to the Google Play Store version so that I could use Chromecast, which they strip from their F-Droid builds because the underlying library isn't open-source and is deemed "impure" by F-Droid (it gets you a "This app has features you might not like" banner, when honestly it's a feature I specifically want).
A similar thing is true of Tempo (a Subsonic-client music player), where the F-Droid builds have Chromecast support stripped out, but the GitHub-published builds have it (so I also have to install Obtanium to get those updated).
"I want to listen to my audio on my devices in my house" is a weird thing to exclude in the name of open-source purity.
Otherwise, I love F-Droid. I just wish they had a bit more nuance to recognize that Chromecast support isn't the same as "constantly reports your location in the background to corporate servers", and so those shouldn't have the same severity of warning banners applied.
>We don't think we have the competence to be a teacher. We would never presume to teach someone else's kids.
While this is a good and rational awareness of one's own capabilities, as someone who grew up in Bible-belt homeschooling circles and saw a wide variance in approaches and effectiveness, the "homeschool co-op"/"homeschool group" model where one parent teaches one subject to many kids, classroom-style, is super common. See, for example, "Classical Conversations" [1], a pretty common one in my area, that leans on "parent as classroom teacher to many kids", without much in the way of prerequisite qualifications.
The population of New Orleans is less than 10% of the population of Louisiana. Also, I'm not aware of a single metric by which Louisiana is considered to be last place in K-12 education in 2024. State PISA scores are basically on par with California, spending per student is higher than Nevada. It's certainly no Massachusetts, I've seen it ranked around 40th out of 50 on different metrics, but I think you're revealing some ignorance in assuming Louisiana's school system is ranked last in the nation.
Pure economics and rational decision making are the exact reasons for engaging in regulatory capture, bribery, and oligarchy.
Why on earth would democracy (or any other form of shared power) be a rational choice for you, from an economic standpoint, if you already are wealthy enough to neuter it to the point where nearly all profits and decision-making authority are allocated to you?
Dictatorship is the ultimate in rational decision-making for a rational self-interested actor. Philanthropy and benevolence are not rational for the wealthy and powerful.
Income inequality and regulatory capture are features of the free market, not bugs. They are baked in by design.
Most countries in the world "patch" those bugs by regulation that moves them away from being pure "free market" economies. Antitrust regulation is a well-known example of this.
To play the devil's advocate here, as someone who grew up homeschooled and in a culture of "micro-scale evolution exists, but macro-scale evolution has not been demonstrated":
>Antibiotic resistance
...is a micro-scale adaptation, like an organism's immune response. Recognizing it does not require belief in a prehistoric common ancestor for all organisms; it just requires observing changes that happen on a much smaller and more rapid scale.
>Existence of vestigial structures in organism. Why do humans get goosebumps when we don't have enough hair to insulate us? Because it's an evolutionary leftover from our hairy ancestors when the reflex would actually cause hair to trap more air for better insulation.
This is non-falsifiable conjecture about a pre-historic past based on observation of present structures. It is equivalent to "we obviously know that dinosaurs did not have feathers, because their skeletons do not have feathers, and feathers would have made them more visible to predators, so they wouldn't have had feathers."
>Understanding evolution is crucial for crop management. The development of pesticide resistance in insects follows the same principles as antibiotic resistance.
...which, again, is a micro-scale adaptation, like an organism's immune response. You can notice pesticide resistance occurring in pests and rotate your pesticides without having to sign on to the unverifiable claim that this happens because all life derives from a single organism.
>Medical research often relies on animal models because of shared evolutionary history. Our biological similarities with other mammals exist because of common ancestry. Without this framework, it becomes harder to understand why medicines tested on mice or primates might work in humans, or why certain diseases affect multiple species similarly.
This is more non-falsifiable distant-past conjecture based on observation of current structures. Is it necessary to believe a particular set of conjectures about the origins of mammals' biological similarities in order to recognize the fact in front of you that the mammals are biologically similar, and thus some mechanisms of action may apply across species, provided those similarities are retained?
>Human susceptibility to back and knee pain is a consequence of how recent bipedalism is in our evolution. Same for why humans are so prone to chocking, our larynx evolved to enable speech at the cost of making it easier for food to enter it.
...which, again, is non-falsifiable distant-past conjecture that has no bearing on recognizing the existence of the verifiable current-day reality in front of you: humans have back and knee pain. Is it necessary to accept a particular set of unprovable conjectures about the distant-past origins of this particular skeletal structure in order to make decisions about how best to treat a symptom that exists today resulting from the skeletal structure that you see immediately in front of you?
But the micro vs. macro distinction is only one of time and scale and that's the whole point: species aren't "real," even fish aren't "real" in any ontological sense, but the countless organisms that we categorize as such existed, exist, and will continue to exist regardless of how we conceive of them.
The ask of evolution and science in general is to accept the incredibly narrow capacity of human cognition as a starting point for an even deeper understanding rather than an end goal to rationalize towards.
>But the micro vs. macro distinction is only one of time and scale and that's the whole point: species aren't "real," even fish aren't "real" in any ontological sense, but the countless organisms that we categorize as such existed, exist, and will continue to exist regardless of how we conceive of them.
This is an excellent rebuttal to the micro/macro distinction, because it's working in the correct direction, which you've stated well:
>to accept the incredibly narrow capacity of human cognition as a starting point for an even deeper understanding rather than an end goal to rationalize towards.
Using the notion of "species" as a "ground truth", as though it were some biological law, is a self-defeating point precisely because the definition of "species" is "a somewhat-arbitrary taxonomy developed by people to try to group organisms together based on observed common traits."
Want to have your mind blown? The creationist fallacy of "irreducible complexity" isn't just wrong for eyeballs and flagellum but for upward complexity as well. And lateral complexity.
OK, for falsifiable how about evolution predicts patterns of genetic similarity between species that match their apparent morphological relationships - a correlation that didn't have to exist but does.
That's not what falsifiable means. It's not experimentally verifiable. There is no way to conduct a test that would negate it if it were untrue.
It is why, being intellectually honest, the theory of evolution as the origin of species is called a "theory" in the academic sense: it's a proposed model that fits the data available on hand, but which has not been experimentally verified in its premise. Short of time-travel, I'm not sure how it can be experimentally verified.
"Falsifiable" means "I can construct an experiment that could yield an outcome that directly demonstrates this idea as false." This is sort of like the difficulty that exists with the four-color theorem [1]: yes, you can run a lot of examples using computer-assisted proof tech, but at best what that tells you is "we haven't found a counterexample yet."
Except, for non-falsifiable claims like the theory of evolution as the origin of species, there is no experiment you can run to provide a counterexample. The theory covers any possible counterexamples by simply saying "that form of life must have evolved from a different origin point and/or under different conditions (regardless of whether we can recreate those conditions)", and tucks any counterexample in neatly into itself without feeling threatened by falsifiability. It is "total" by having an "escape hatch" for any counterexamples.
That stacks it up alongside "a deity made everything, and designed an ordered universe with certain mechanics, including giving organisms the ability to adapt"; both are explanations that fit the available data, but neither can be experimentally verified. Similarly, that theory is "total" by having an escape hatch: "well, maybe the deity did something different in that case." Young-earth Creationists do this with visible starlight that is a million or more lightyears away: "maybe God just accelerated that starlight so that humans would have a pretty night sky."
That tendency is similar to "maybe the [hypothetical] organisms on Mars adapted from a different common ancestor that maybe was made of non-living substances that are similar to the non-living substances that comprised Earth's first organism." Boom, done, no need to re-examine the premise, you just fold it in with "maybe the same magic worked a little differently over there," just like saying "maybe God made starlight go faster in the direction of Earth."
As long as you don't engage in denial of the available data because of your theory, then I don't understand why holding a particular non-falsifiable theory is mandatory.
It doesn't matter if I hold to the theory that the universe began as an origin-less hypercompressed single point of matter suddenly and rapidly decompressing...if I'm in the lab next to you claiming that vaccines cause autism. The problem is not which non-verifiable theory I hold about an unrelated subject, but rather my denial of the available data on hand.
Similarly, it doesn't matter that Louis Pasteur was a Creationist when discussing the mechanisms he discovered by which vaccines work. What matters is his recognition of the reality of the data at hand, and his work to explore and build on it.
>Some have even suggested “y’all,” a word that reads as much too slangy, regional or what you might even call ethnic to ever gain universal acceptance.
Heaven forbid we embrace something that could be called "ethnic"! Something something must secure a future for our linguistic monoculture.
I wonder if this writer would've written the same thing, but with "bruh" or "fam" swapped in:
>Some have even suggested “bruh,” a word that reads as much too slangy, regional or what you might even call ethnic to ever gain universal acceptance.
Probably not, as that would've been immediately recognized as racist (because "bruh" is ethnic, despite its recent widespread adoption, in the sense that its roots are in black-culture slang – just like how "y'all" is ethnic, despite its recent widespread adoption, in the sense that its roots are in southern-culture slang).
I dunno why the NYT author feels the need to find ways to be linguistically xenophobic.
Seconded. In my entire North Carolina + Georgia upbringing, with frequent visits to Tennessee and back to NC, and some visits to South Carolina, Florida, and Alabama...never once have I heard y'all referring to a single person.
It's literally a contraction for "you all". That doesn't make sense for talking to one person.
Poetry, like humor, involves the use of the reader's expectations, but is typically most effective when subverting those expectations.
There's a lot of bad poetry in the world that just follows readers' expectations. I should know, I've written some of it. Unfortunately, I'd suspect that most readers' understanding of poetry lacks that crucial element of subversion, and so an LLM – which mostly just spits out the most-probable, most-expected next token – looks like what people think poetry is.
An LLM would not have created the curtal sonnet form, because it would've been too busy following the rules of a Shakespearean or Petrarchan sonnet. Similarly, an LLM wouldn't create something that intentionally breaks the pattern of a form in order to convey a sense of brokenness or out-of-place-ness, because it's too locked in on the most-common-denominator from previous inputs. And yet, many of the most powerful works I've read are the ones that can convey a disjointed feeling on purpose – something an LLM is specifically geared not to do.
Poetry aims for the heart, and catches the mind along the way. An LLM does not have the requisite complex emotional intelligence, and is indeed a pretty poor simulation of emotional intelligence.
Consider Auden's Epitaph on a Tyrant, which is powerful because it is so suddenly shocking, as it describes something that sounds perhaps like an artist or author, until it very suddenly doesn't, on the last line:
Perfection, of a kind, was what he was after,
and the poetry he invented was easy to understand;
he knew human folly like the back of his hand,
and was greatly interested in armies and fleets;
when he laughed, respectable senators burst with laughter,
and when he cried the little children died in the streets.
One could literally take Claude 3.5 Sonnet New or o1-preview and disprove this in an hour or two just by prompting the AI to try to exhibit the type of poetry you want and then maybe asking it to do a little bit of automated critique and refinement.
You can also experiment with having a higher temperature, (maybe just for the first draft).
You claim that LLMs can't make poetry like that. I bet they can if you just ask them to.
They could, but they probably won't. Poems like GP are basically using the power of emotional manipulation for good, and companies like Anthropic try very hard to prevent Claude from having that capability.
What he said was basically that it just couldn't create unexpected verses or break form. Since supposedly it can only do the most probable token -- but that's not how sampling works unless you use temperature 0. And it can easily be instructed to break from a strict form (which would create a new variation of the form) for effect if it made sense.
You could also ask it to create a new form and it could. I don't work for you so I don't have to create examples, but anyone who has used the latest SOTA models for any amount of time knows this capability is expected, and if you were really interested then you would try it. If you feel the result isn't very good, ask it to improve it.
I could program even a markov chain to generate a lot of odd unusual potentially interesting stuff, but no one would call any of it a new form of poetry, because establishing something like that requires social status, which robots don't have.
It can be obvious and go to the heart. I'm not sure Wilfred Owen's Dulce et decorum est is anything other than straight down the line, but it made me cry when I first read it.
That said, maybe the subversion is in how the reality is contrasted with the marketing.
I see 'subversion' as more broad. In good poetry, subversion is constantly happening at a micro level, through playing with meaning, meter, word choice. I think it's very easy to identify AI-generated poetry because it lacks any of that -- but on the flip side, if you don't understand the rules, you don't understand how to subvert them.
Even in Dulce et decorum est -- though the meaning's straightforward, there are plenty of small, subversive (and pretty) ideas in the meter. For example, the line "He plunges at me, guttering, choking, drowning" is an unexpected, disruptive staccato that really sounds like guttering, choking, drowning. It's a beautiful poem and is overflowing with such examples.
(I think this applies to art as a whole, not just poetry.)
Even the last line of the poem is a twist on expectations and plays on the irony of nationalism and patriotism, the beautiful latin phrase that inspires young men and the harsh reality of trench warfare that it leads them into. No LLM is going to do that.
Funnily enough, the phrase doesn't even shy away from the fact that you are dying for your country. The poem really gets at the fact that the way you die is horrible and that in itself is enough to counter the romantic notion. The reality of dying for your fatherland is not cleanly and painlessly dropping from your mount in a heroic and successful cavalry charge, even if that is good.
>There's a lot of bad poetry in the world that just follows readers' expectations. I should know, I've written some of it.
Apparently, in 'the good old days (of the internet)', your poetry would be published by yourself, on your webpage - complete with a starry-twinkling background, a 'number of visitors' counter in the bottom right, and in a pink font.
type 'a Tree =
| Node of ('a Tree, 'a Tree)
| Leaf of 'a
Is saying "The possible cases for the type describing an 'a Tree are: a Node of two other 'a subtrees, or else a Leaf of an 'a."
Then when you concrete-ize it, you have an "int Tree" or a "string Tree".
This is a little clearer with thinking of the List datatype: this is a number list, that's a string list, that's a grocery list, this is a guest list...etc.
One nice thing in F# is that there's an ability to "standardize" how you write generics to look more like the (admittedly arbitrary) way that C# or Java or C++ write them:
type Tree<'a> =
| Node of (Tree<'a>, Tree<'a>)
| Leaf of 'a
> This is a little clearer with thinking of the List datatype: this is a number list, that's a string list, that's a grocery list, this is a guest list...etc.
That's what I was referring to when I said "as a non-native english speaker". In my native language I would say those in a way that's more similar to "a list of numbers", etc etc.
> One nice thing in F# is that there's an ability to "standardize" how you write generics to look more like the (admittedly arbitrary) way that C# or Java or C++ write them:
Arguably this order (type then generic arguments) is not that arbitrary because this is effectively a type level function, so it makes sense that it uses the same order. However the angle brackets are indeed arbitrary, and IMO to be more consistent the same syntax as function application should be used. For practical purposes though this may make a type's syntax too similar to an expression, so it may not be the best choice (unless you're in a language where types are actually values)
A similar thing is true of Tempo (a Subsonic-client music player), where the F-Droid builds have Chromecast support stripped out, but the GitHub-published builds have it (so I also have to install Obtanium to get those updated).
"I want to listen to my audio on my devices in my house" is a weird thing to exclude in the name of open-source purity.
Otherwise, I love F-Droid. I just wish they had a bit more nuance to recognize that Chromecast support isn't the same as "constantly reports your location in the background to corporate servers", and so those shouldn't have the same severity of warning banners applied.
reply