Damn, I was afraid this was going to be someone stealing my genius idea. Gladly it was just tangential to that idea.
Idea: Train an ai to construct entire threads based on the input of the title and link. There should be plenty of training material, and frankly I think it would be hilariously similar to the real threads.
When new submissions are added to HN, the system will fetch the title and link, and predict the entire thread. You can view the thread on something like hnpredicted.com/?id=33680661 . When the real thread has had no new comment added for five days, it is reintegrated into the system further training the model.
Possibly, the model could be trained on the actual content of the link. But I suspect just using the short title might be better. Larger concentration on the actual material there, websites contain lots of text which is useless.
Now that I've posted the idea, feel free to steal it ofcourse. It's documented that I came up with it anyway. :)
That’s what this site is. You’ve been talking to a machine all this time. Do you really think a person would have this much time to argue day in and day out over systemd? That that many people care wherever or not Go has generics and which other languages does them best? That rust somehow is the cure for everything from game engines to the common cold? Think about all the times a “person” stated something was Turing Complete and how it made zero sense to you. It’s been machines all this time.
Yeah, but at some point those machines would have had to have original training data, so at some point people probably did have time to engage in techno religious wars. Unless of course, all of this is made up, to keep us occupied to not discover the matrix or alike ..
> Possibly, the model could be trained on the actual content of the link. But I suspect just using the short title might be better. Larger concentration on the actual material there, websites contain lots of text which is useless.
Most people do not read the article before commenting, you do not have to train the AI to do that.
There was a reddit experiment a few years back, SubredditSimulator, that was a scaled version of what you described. It was obviously more surreal and funny than accurate. There were users trained off of popular subreddits, and they just ran with threads and comments being created purely from the data.
I did this quite a few years ago. If I remember correctly, it was so convincing that people emailed Dang because I set the username at the top to be “pg”
Headers full of navigation. Sign up for my newsletter spam. Footers. The webpage isn't just "the document we're talking about", but an entire application most of the time.
>this is a fully-automated subreddit that generates random submissions and comments using markov chains (see below for more info), with each bot account creating text based on comments from a different subreddit.
I present algorithm whereby each byte is iterated from 0 to 255, following recursively for next byte for each previously yielded result. From above nonsensical results are removed.
Above algorithm generalizes all human knowledge ever conceived, in any format, as well as any future (duplicated!) ideas that may be conceived.
I donate above algorithm and all its results to public domain.
> I present algorithm whereby each byte is iterated from 0 to 255
You should've just said a stream of bits, as your system cannot generate anything that requires a non-multiple of 8 bits to be accurately expressed. Zero padding is not the same as absence of bits.
I'm also surprised someone hasn't made a parody LLM-based HN thread generator yet. Perhaps until recently they were too busy writing crypto startup business plan generators to bilk gullible VCs.
Let's go a level further. Instead of banning/shadowbanning people just send it to shadow realm where all they see are bot created comments and bot created answers to what they comment
Some video games already did something like that, except instead of AI they put you in match with other offenders, in sort of personal hell, then only let you out once you won some matches
I love this idea—I think it would be fun to gamify it! Keep the predicted thread secret, and then use it to award points to hn commenters based on originality. If a comment is functionally the same as a predicted comment, it gets -10 points. If the comment is novel, it gets 10 points. Then there's a sliding scale between.
You'd need some checks to prevent people from posting a bunch of Zalgo text (maybe a model to identify "interesting" or "insightful" posts), but it could be a fun overlay on the hn commenting ecosystem.
You are welcome! My very small contribution to the system (I asked and "pushed" for it) - of course, Dang and/or his team implemented it. It is pleasant to read about it in such appreciative terms.
[-] collapses the replies so that you can easily see what's next. On reddit the equivalent functionality is a button literally called "hide child comments."
But as another comment mentioned, you can click root | parent | prev | next to get navigated through the thread. Personally, I prefer the collapse, though, as having the top of my screen move without my direct control is a bit jarring, but there's more functionality from these buttons.
Important fact:
The [-] does not remove a point. After you get 500 points, you earn the privilege of down voting, you will get a "down triangle" underneath the "up vote up triangle" that you can click on.
Or about how React bloatware introduces tens of petabytes of completely unnecessary dependencies, is destroying the Web, a perfectly-designed hypertext platform for linking documents together for reading on all manner of devices from screenreaders to Braille interfaces and toaster ovens.
How does this compare to X? (where X is an obscure project on github with no documentation of what it's feature boundaries are, has at best 1000 stars and further internet sleuthing suggests its at best only tangentially associated with the original comments domain.
The end result being the author feels obliged to figure it out and reply incase not doing so seems arrogant and/or poorly educated in the domain)
Off-topic aside, but why do people complain about back-button hijack when you can long-press or right-click and get out of the site? It works on any browser AIUI.
1. Long press doesn't show full history, it limits to 10-15(? Might be more). I've seen the occasional scummy background script that exploits this with numerous scripted redirects (usually pages with seo snippets cloned from SO). You have to disable the network to go back without being re-redirected, then go through a couple rounds of the back button history to get back to your original page.
2. If you're not using a full browser or are set up to keep a short history, you're basically SOL (Materialistic for HN has a built-in WebView, and I've used similar apps for other sites)
3. I shouldn't have to use history just because some site decided that it should add multiple history entries when I tried to scroll down an infinite list. I'm only on your website because it showed up in a search result, I'm not here to browse other unrelated content.
As a US poster, I need all units of measure to be what the Founding Fathers used and all pickup trucks to be what the Founding Fathers drove.
As an EU poster, I won't understand the US obsession with large vehicles and I will recommend bicycling instead.
As a programmer, I think we can improve performance if we multithread, containerise, microservice, and store n-depth JSON in a No-SQL database.
As a Python programmer, this doesn't look Pythonic. We should be using pythonic Python, especially the new release that adds the syntactic sugar that we've all been waiting for.
As a person running a single static website, I can confidently say that Kubernetes is the most overhyped technology ever. After having eventually realised that my tediously simple workload can be deployed effectively using a short systemd unit file and an even shorter ansible script, I no longer believe anyone actually runs production workloads on Kubernetes.
I will now be considering other options for the service mesh that my startup clearly needs to implement in order to send an occasional email newsletter.
My start-up has rewritten Kubernetes in Rust because Go is slow and doesn't implement traits in a rustaceous way. Now we have a system about which we still can't reason tractably, but we can guarantee memory safety wherever memory safety can be guaranteed.
What nanny-state whiners don't understand is that helmets can actually twist your neck or choke you as you're flipping in the air and about to hit the pavement head-first. But it's worse than that, helmets give people a false sense of confidence. That's the real problem with our civilisation, too much confidence that death isn't imminent at every turn. Forewarned is forearmed, I always say to my shrinking coworkers. Oh yeah, pry my weapon from my cold, dead fingers.
As a(As a(As a(recursive defined class of person)))) im astounded by how important perspective and caste has become for some, while we are all just the same hairless monkeys
"As an American poster, I will get really upset about the idea of GDPR applying to US companies that do business with EU citizens and businesses. I will also complain about other EU regulations, because regulation is bad, stifles innovation and we don't need it in the land of the free, unless the state is California."
"as an American poster I will also get annoyed by non-americans having anything to say about how our legislation works and how that affects them; after if it wasn't for us there wouldn't be any money around"
> Here's a long detailed, objective explanation of everything related to this issue. It's probably more useful than the actual link and it may serve as one of the best efforts to consolidate information on this subject on the entire Internet. If it contains original research only a couple of readers will be qualified to tell. Half the people who upvote this won't understand more than the first two paragraphs.
Exactly explains the value of hn, when this happens.
There are so many problems that affect many working programmers which you only learn the good answers to by lucking into spending years in the right job.
I always upvote when people share that specialized knowledge and edify the rest of us.
Absolutely. Sometimes you get a comment thread where some specialists add way more than “average” commenters could actually add. I like it because it gives insights in difficult subjects!
I disagree with the author. I know he's incredibly successful and right about pretty much everything he's ever said, but I've had some experience in this area and just finished reading through some of the archives and I think his focus is wrong. I'm going to ignore the technical issue and talk about the bigger picture and higher level things than what was said in the blog post. If the OP thinks that the process is most important, it's really about end results. But if he thinks it should be about the end results then he's an idiot for not thinking about the process. I'll weasel in a reference the startup I co-founded even though it's not directly relevant.
Yes, I can elaborate, however, this information should have been taught to you by the age of 2. Silly HN.
A tangent is simply a line that touches a non-linear curve (like a circle) at only a single point. It represents an equation with the relationship between the coordinates “x” and “y” on a two-dimensional graph.
The tangential velocity is the measurement of the speed at any point tangent to a rotating wheel in a circular motion. Thus angular velocity, ω, is related to the tangential velocity, Vt, through the formula. Tangential velocity is the component of the motion along the edge of a circle measured at any arbitrary point of time. As per its name, tangential velocity describes the motion of an object along the edge of the circle, whose direction at any given point on the circle is always along the tangent to that point.
The only thing missing is the actual expert making a clueful statement at the very bottom of the page where nobody reads. Possibly downvoted a couple of times.
This is cool (only read the first two lines so far) but if you're really interested in Hacker News Parody software you should check out our startup: https://zombo.com/#HACKER_NEWS_PARODY_DOT_IO
But to the ultra self important pedants, they don’t see it that way at all…it’s not useless it’s a fundamental reality that you’d be a complete idiot not to fully appreciate…just make sure you also worship their brilliance in deigning to illuminate you
It's missing a top-level post which makes an analogy, which turns into a ten-deep thread of people arguing entirely about the analogy, the friendly article having been completely forgotten.
I posted before on HN about this trend, of commenters often completely forgetting about the original article, and arguing a lot (in child comments) about some top-voted comment.
I called it the Hacker News Right-Side Effect (or Rule)
It’s missing a comment from someone who dismisses someone else’s comment by starting with “Ehhhh no.” It’s also missing someone making a funny joke and then being downvoted because HN comments need to be serious contributions to the topic.
A couple years back someone posted something I wrote. I’d written it in November 2018 and it was currently March 2019. Someone started a thread that it needed (2018) and the HN mods actually made it happen - as well as changing the title.
In all seriousness, I think people not reading the full article is a recurring theme on hn.
For controversial topics, it's a major problem: people don't establish what's actually being debated, instead attacking/defended what they think I'd being debated, so the conversation develops slowly as a result.
However, I also think it's a strength of hn, because hn has a strong culture of going off topic: allowing for people with related experience to share and to suggest new avenues to explore. Often, instead of bookmarking the referenced article, I bookmark the hn thread instead for this reason.
Not enough nitpicky corrections that devolve into even more nitpicky nerd arguments about semantics and usage of words that are entirely missing the high-level picture or the ability to understand the meaning of sentences based on their context.
Also, this could use more critical replies discussing why this project couldn't possibly work, or how this article is a result of living in a bubble.
And, of course, someone managing to steer a technical subject into a divisive political argument.
I'm someone who clearly didn't read the article, because I raise several points which were directly addressed by the author, but I do not refer to them at all in my very angry comment. It's possible the only thing I know about this article is the title, which it's also possible I misread in my eagerness to have my thoughts known by the internet.
This is pretty good, but to be peak HN the thread needs to have 'showdead' on and include three greyed-out comments: One claiming that the OP is wrong because he is a nefarious Jew, one half-page comment of entirely incomprehensible word salad, and a comment trying to sell boner pills in broken English with a misspelled URL that goes nowhere.
Too bad that n-gate.com stopped updating in mid-2021. It was great for this stuff, and arguably it still is.
Edit: Don't miss the usernames in OP's parody thread though, they're just as hard to spot as on actual HN (which is a great feature of this site) but I thought they were the funniest part.
When telling people about HN, I'd emphasize first reading the guidelines, and not treating it like other social media. Then I'd say, when some comment thread goes intolerable, they can get relief by reading some n-gate. (Though the content of the last post precluded mentioning it at the company where I was working.)
I hope the n-gate person is well, and that they didn't blow a gasket, taking on the burdens of too many HNers.
This is missing a tangent where someone finds a way to interject a point against “woke,” gets mostly downvoted into gray text, and is followed by a ten level deep rehashed debate over whether free speech should imply access to private platforms or not. Oh, and the various dead troll replies that you can see if show dead is on.
> Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage. They're too common to be interesting.
On a tangential note, did this guideline change recently? In some search results "e.g." changes to "things like".
While I like the idea in general but instead of hosting it as a static file, SQLie table would have been a more accessible approach not to mention that it is far easier to query from Rust.
Heck you could even compile SQLite and Rust both to Webassbly as Cloud flare edge workers but I'll not go that far.
1. I noticed a word used slightly incorrectly in the title. Rather than reading it with the clear intended meaning, let’s take it as literally as possible to build a suitable straw man.
2. Someone replying in every thread, debating everyone like it’s their job and accounting for half the replies in the topic.
A serious post (I know, it's hard to tell at this point, isn't it?):
Why does HN need the karma points system at all? Downvoting is hardly used on other social networking sites any more, which should tell you something. Getting one pedantipoint has to be a motivation for "I have stuff to say about JIT." posts.
Blocking & muting people is quite effective on Twitter, for example. The dream of a civil discourse with people disagreeing politely was obsolete before the Web was invented. There are a lot of accounts which someone here would rather never see, so why not just let those people erase them, just for their own TL?
Automatically downvoting anything you don't agree with is, basically, "Lord of the Flies" behavior. It leads to brigading.
HN “needs” karma points because HN has karma points and is over all a good forum. Changing it at this point risks changing the balance that makes HN what it is.
I don’t block or mute people on forums where that is allowed because I don’t want to see a different conversation from everybody else.
Don't mind me, I'm just going to go around this parody thread downvoting anyone who talks about their small bootstrapped passion projects, accusing small funded companies that people like of astroturfing, and then storming off in a huff when I get called out. ;)
irt "...I know this is off-topic, but does anyone know how he got visual effect X on his blog? ..."
I cannot describe the sadness I feel when I see the title of an essay I might enjoy, decide to use HN comments as a way to decide, then the top comments are all "Why is he using comic sans as a font? Does he hate us?" and so forth.
A kitten dies each time this happens.
Over time, I have participated less and less on HN, but I still use it daily to screen out new tech news. This site rocks. Excellent parody!
Also, for the first seventeen people to sign up for my free e-book, I will purchase certificates for free rides from Uber For Horses. Even if you don't like the book, think of the horses!
Not enough sophisticated adjactives. On HN everything is incredible, amazing, electrifying and goosebumps inducing. I would love to see a frequency analysis of such words on HN vs elsewhere.
You missed the mandatory post on how I can read the author's blog post for 400 hours without recharging on my superior Mac M1, versus just 10 minutes on pleb Linux boxes.
You don’t have to continue the parody and prove that the OP is right by making more parody comments. We already know he’s right. There has to be a word for that.
They missed the commentor who has nothing substantial too add too the actual conversation but managed to hone in on a piece of English semantics that needs to be endlessly nitpicked.
Inevitable followup "reddit level" comment left as an exercise for the reader.
I wish I could, in the true spirit of NH, down vote every comment I don't agree with on this tread. Unfortunately, that would really count against karma. But it's tempting :)
That said, the parody shows nothing down voted. Is that part of the parody? Or an oversight?
Needs a “person that posts an entirely reasonable analogy” followed by a bunch of posts arguing whether the analogy covers some small niche edge cases, with at least one person quoting a definition from a dictionary.
It misses the <template> guy, posting about his life long obsession in every thread, no matter how unrelated. Also no dang, providing previous discussion links and stomping out flamewars caused by the usual suspects.
One thing it's missing is a reply to the comment saying "have you read the article" asking whether they have "read the Hacker News rules" as it forbids asking if someone has read the article.
I don't find this parody funny at all because the site doesn't work without JS plus it's taking over my scrollbar. Not at all smooth in Firefox 3.11 with everything except HTML disabled.
This is rather redundant, as HN often manages to parody itself. Does that make the intentional parody, being almost indistinguishable from real HN, a good or a bad one? I'm torn.
The URL should be to an article from 5 years ago with the OP saying that he’s sure nobody has seen this before, and 15 replies with links to all the times that URL was mentioned on HN in the past.
I almost upvoted some fake comments there. Is that bad? Am I conditioned to see the world through an orange and grey filter, letting only sarcasm and earnest pedantry through?
Here would be a long winded thread debating a very pedantic point while using the definition of different fallacies to argue the other poster makes no valid points.
An internet attempts a third-rate mockery of Hackernews. Hackernews pulls a "this, but unironically", earnestly replicating every joke comment in the original article, right down to the obligatory XKCD.
This is missing the “scrolling is broken, history is broken, text isn’t high contrast enough, doesn’t work on mobile Safari, I hate paywalls, I hate cookie pop-ups, I hate newsletter signups, doesn’t work with adblock, doesn’t work without JavaScript, so didn’t read the article” comment.
It’s missing the part where someone suggests stupidity and obtuseness on the part of anyone who questions free markets outside of theoretical experiments.
Similarly, it’s missing the comment from a woman-identified person that is followed by negging from male users.
Idea: Train an ai to construct entire threads based on the input of the title and link. There should be plenty of training material, and frankly I think it would be hilariously similar to the real threads.
When new submissions are added to HN, the system will fetch the title and link, and predict the entire thread. You can view the thread on something like hnpredicted.com/?id=33680661 . When the real thread has had no new comment added for five days, it is reintegrated into the system further training the model.
Possibly, the model could be trained on the actual content of the link. But I suspect just using the short title might be better. Larger concentration on the actual material there, websites contain lots of text which is useless.
Now that I've posted the idea, feel free to steal it ofcourse. It's documented that I came up with it anyway. :)