The AI startup drama that's damaging Y Combinator's reputation
indiehackers.com
indiehackers.com
Is this true? I never thought HN moderated content critical of itself
That article was of the type that we'd normally downweight off the front page, since it wasn't substantive or intellectually interesting in the way that the site mandate calls for. Since it was about YC, though, we cut the amount of downweight. The post went from #1 to #15 instead of (say) #55 or whatever. #15 on HN's front page is far from "disappeared". It's still a prominent position, and the post continued to receive a lot of attention and traffic that it otherwise would not have.
> What your explanation is hinting at, although you're not saying it outright, is that "Yes, that's true."
I wouldn't say so. The reason for downweighting the post was that it wasn't a good article for HN. The reason for downweighting the post less was that it was about YC. You can call that "deliberately lowered in the rankings" but it's equally true to say "deliberately heightened in the rankings". It depends on which baseline you're comparing to. Relative to the baseline of standard practice on HN, the post was heightened, not lowered.
Is that clearer? I'm not asking you (or anyone) to agree with, like, or approve of how we moderate HN but I would like the details to be understood.
Dang said they didn't do it and it's being flagged by users.
Which obviously makes sense of you think about it.
"Company investigated Company and found that Company did nothing wrong"
"dang" is a detail--the point is, if someone is being cut a paycheck by a company, the public is well within reason to believe that person has a job obligation to favorably represent the interests of that company.
Would anyone stop using HN if you knew they buried negative stories about YC? You're all still here so likely no.
Explicitly saying you don't do X while doing X is lying, whether or not you are getting a paycheck for it.
(And to be clear, I doubt dang is lying; there’s no need to resort to centralized moderation to explain the observed behavior.)
Yes, but the way we do it is different from the common assumption because we want to optimize HN's value for YC globally rather than overreacting to any particular story.
HN's value to YC consists of the community, and the community only exists because of goodwill and trust. To jeopardize that for the sake of suppressing a particular story (even if the article is false and/or sensational and/or shallow, silly or whatever) would be a super dumb tradeoff, so (as my son once said when he was little) "that what we not do".
I understand the skepticism, and of course you're free to disbelieve any part of what I say. All I can do is explain to people what we do and how we think about it, answer questions when asked, and hope that this is good enough to keep the bulk of the community happy.
in other news: man kills his family, neighbours swear they were a normal happy family, It's totally unexpected, we never would have thought something like this could happen
past behavior is a good predictor of future behavior if and only if nothing else changes.
not speaking about dang specifically, but IMO it's a bit different to not do something when it benefits your reputation and to not do anything when it can harm your job safety.
https://hnrankings.info/41697032/
this happens very, very frequently
I understand that articles don’t just climb due to the magnitude of upvotes, but also the velocity or upvote rate.
All kinds of articles don’t stick to the top, to me the more likely explanation is that the rate of upvotes was not sustained.
With both fewer upvotes and comments.
https://hnrankings.info/41721668/
8 of the current top 10 stories have been on the front page for longer than the submission that is critical of ycombinator and every single one of them has vastly fewer upvotes and comments. It's not even close, the story that is currently in position 6 has 1/4th the upvotes and 1/10th the comments. It has been on the front page since its submission.
https://hnrankings.info/41721318/
It's been on the front page for 11 hours.
It is impossible for the velocity, given any reasonable common sense examination, for the upvotes on that post to be greater than the story that was nuked after 2 hours.
There is no submission on the front page, some of which have been on the front page for over 24 hours, that has more upvotes, comments, or any conceivable rate of upvoting or commenting that even approaches 1/10th of the nuked story.
There is one submission that has an average of four upvotes per hour.
Assuming that upvotes fall off precipitously after leaving the front page, which I would say is a safe assumption, the nuked story had an upvote rate of several hundred per hour.
There's something fishy going on and that smell isn't the strong odor given off by Salt Water Dimmers, a submission to a barren wikipedia page about an obsolete technology with 13 upvotes and 5 comments that debuted on the front page, and has been there for several hours.
I'm not joking. On the front page of HN for several hours is a link to a 200-word wikipedia article about dimmers used in stage productions: https://news.ycombinator.com/item?id=41687950
For what it's worth, stories about Boom (a ycombinator joint) SEEM to get nuked extremely rapidly when non-VC non-techbro domain experts start chiming in about what their chances for success actually are and how there's a 50% chance they're the next OceanGate and a 50% chance they're just a scam that got way too big for its britches.
Like the CA admissions thread wasn't on the frontpage initially for under 24h and it had 3x the comments as the YC thread. https://hnrankings.info/41697032,41700516/
Community goodwill is the only value HN has, so we take our lumps when we have to, rather than jeopardize that.
"Less" doesn't mean "we don't moderate at all"—that would be too big a loophole. "Less" means "do what we normally would, but not as much". That way we can keep the front page reasonably close to the site mandate while still having a consistent approach to conflicts of interest.
For example, if a story is the kind of thing we'd normally downweight off the front page (e.g. because it's a typical opinion piece or drama that isn't intellectually interesting), then "do what we normally would, only less" might mean that the article ends up halfway down the frontpage, whereas normally we'd downrank it off the frontpage altogether.
This approach goes back to the first morning that pg was showing me how he moderated HN and it was literally the first thing he said to me, before I had a chance to grab a chair. He kind of barked it actually - 'whatever you do, don't do that!'
10 years later, it has held up well: it's a simple rule, easy to be both transparent and consistent about, that addresses one of the hardest aspects of running a site like HN. It doesn't work perfectly (nothing on HN can work perfectly, for the simple reason that different segments of the community want different things) but I find it hard to imagine a better tradeoff.
This doesn't stop people from jumping to inaccurate conclusions (such as "HN mods suppress bad stories about YC" when in fact we do the opposite), but it does mean we can answer questions in good conscience, which is vital not only to community goodwill but also our own morale.
As with anything you have "the powerful people" who have opinions on what critiques are "fair", and will demand them communicated with perfect decorum.
Additionally, people are aware of how it will "look". In a way they don't care, but in reality there's always an appearance people prefer to keep.
Anyway actually break out some pointy critiques, and they'll get mad.
> dawg i chatgpt'd the license, anyone is free to use our app for free for whatever they want. if there's a problem with the license just lmk i'll change it. we busy building rn can't be bothered with legal
[https://x.com/anothercohen/status/1840515897804623882/photo/...]
As an investor I'd probably pick passion and dedication over maturity and responsibility too
The header image of his X account is also telling regarding his priorities: https://x.com/CodeFryingPan/header_photo
I genuinely don't understand how you can make a business based on open source and not get IP lawyers involved.
If you are depending on an open source project for your startup, why say “Get dat shit outta here” after you change its name to your product? If its shit then why use it?
Our software changed it to the canonical URL it got from that page, which was https://www.indiehackers.com/starting-up/the-ai-startup-dram...
I'll change to your link above now. Thanks!
p.s. I would have done this sooner but wasn't aware of this thread until an hour or so ago
Knowing the title is in reference to Pear (and not something that could be _actually_ damaging to YC's rep) lets me know the article is probably isnt worth the time.
YC's main value is in subsequent fundraising, wherein companies are pre-vetted by YC before being invested in by VCs. If they lose the confidence of VCs as being a reliable arbiter of preseed startups, the better startups will just go elsewhere (already happening) and soon the VCs will too. Thus harming YC's reputation massively.
Including Void, Continue, and PearAI.
YC criticized for backing AI startup that simply cloned another AI startup - https://news.ycombinator.com/item?id=41707495 - Oct 2024 (256 comments)
Pear AI founder: We made two big mistakes - https://news.ycombinator.com/item?id=41701265 - Sept 2024 (228 comments)
Y Combinator Traded Prestige for Growth - https://news.ycombinator.com/item?id=41697032 - Sept 2024 (244 comments)