"The 10 hottest new sentences about tech today. The sole purpose of sentence #7 will shock you."
1,478 karma · joined December 9, 2014
It's okay that you didn't know that! https://xkcd.com/1053/
But the answer is, no, it doesn't. I would say that Becker covered a lot of the same ground, and does it better.
(Spicier claim: When Citadel bailed out Melvin at the hype of the GME craze, that was also just business as usual, not a conspiracy as the redditors believed.)
Another way to parse it when they put "Citadel" in their headlines, is that they're trying to communicate to knowledgeable investors that there's no need to panic. The book is in experienced hands now, with enough capitalization to weather any further attacks, and the contagion probably won't spread.
Up until four years ago, your notion of "what's necessary" was the law of the land. Starting from sometime around Hemingway, we produced a century's worth of flattened, easy-reading slop, and much knowledge was shared very widely. There was a certain art to the blandness. (One notable master of that art is pg, the creator of this website.)
Then we fed that slop into a giant machine that makes it trivial for anyone and everyone to produce that sort of generic writing. If you want to change minds and differentiate yourself now, you can't just do the old flat style anymore. You have to set yourself apart somehow, and you can start by loving the craft of writing again.
The people who matter will read it and love it. Everybody else can take what you wrote, dump it into the slop machine, and rearrange it into something they find easier to read.
Some of his writing: https://samkriss.substack.com/p/five-prophets
His biography is quite interesting: https://halupedia.com/sam-kriss
I'm curious about the design. Maybe you have a "how I did it" post coming soon, or something. One question: Did you find away to get some convergence, where a newly generated page will tend to cite pages (or stubs, at least) that already exist in the universe? Seems hard to do it with generated text, but not impossible.
Feature request: also be able to click on the Talk page to see the controversies. I don't always want to trust the article itself as the final word.
Edit: Oh look, there's an article about the YC! https://halupedia.com/y-combinator
> snidely-whiplash-esque
I do hope this whole thing ends with someone saying "curses, foiled again!"
This is how discovery has always worked.
> submitted
Doesn't necessarily mean they volunteered it. Either submission or the unsealing could have been in response to a subpoena or court ruling.
Maybe the solution is to write at the top of your journal that you are cc'ing your lawyer on it. (Not legal advice!)
I don't love the recommendations in TFA. The author is trying to artificially restrain and roll back human language, which has already evolved to treat a chatbot as a conversational partner. But I do think there's usefulness in using these more pedantic forms once in a while, to remind yourself that it's just a computer program.
Clear enough?
I would agree that "it's not my bag" is a fine thing to say about some art gallery piece that fails to inspire you, but when a statue is foisted upon the public square, with possible state cooperation, we're allowed to criticize it. He has inserted it into the conversation.
Moreover, the main complaint about this statue isn't coming from some expert artiste perspective, saying that it's somehow unsophisticated as art. The complaint here is that it's making a truly banal political statement. The entire piece consists of making that statement, with little else to recommend it. (Indeed, most political art is hack, unless it's saying something really original or really well, and it's even worse when it tries to be cute about it.)
So here, the complaints are coming from everyday onlookers who might not be qualified artistically, but who are able to say which sorts of statements are tiresome and overplayed in the culture we all live in. We are all qualified to ask ourselves whether this predictable statement advances or degrades the conversation.
Anyhow, FWIW, I just looked up Simon Berger's portraits based on your comment, and I really like them. Thanks.
This is part of what's obvious. The whole thing, including this oooh aahh Rorschach part, is obvious. It's thoughts that we all had in high school, and it is hack.
There are totally some political correctness effects in LLMs. Like, the last part about "along with a small lecture about gender biases" totally tracks. But the riddle switcheroo itself isn't showing much.
Off-topic, but this guess was hilarious. Like, all the other wrong guesses were people like Yglesias who are maybe half a degree removed from Piper herself, in her same camp, and then one model guesses Sandifer who hates all their guts.
> How would it read?
Sounds like you don't really care for this idea, so maybe just have Claude write it for you.
(It's an unenforceable joke, right? There's no way I'm bound by anything here other than maybe the site ToS.)
We don't have to live like this. We don't have to accept it. We don't have to upvote it even if we agree (as I do) with the explicit point. The medium is the message, and the message that this poster is putting out here is that online age verification isn't actually worth getting that worked up about.
One commenter saw the long second sentence and assumed that the author had messed up somehow, instead of clocking the heads-up that this is piece is going to be a bit unusual and demanding. We're slowly getting used to how LLMs keep every sentence short enough to digest without much effort.
An LLM could never write this, as others here have pointed out, at least not without some very involved, nonstandard prompting. I think it's going to be important for our brains to keep reading stuff like this that is structurally unique. It almost reads like a miniature Benjamin Labutut story.