Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Reddit is OpenAI’s moat

cyberdemon.org

Post reply on HN
Print view View original

301–310 of 319 posts

Re: Reddit is OpenAI’s moat

#301

Post by pixl97 »

matteoraso wrote: post #293
pixl97 wrote: post #283

Earlier quoted context omitted.

>the opinions i run into real life can be very different then with people in the real world. I think you messed that sentence up, but I get what you're trying to say.... And I think it's this. The opinions you get from people 'IRL' are not apt to be as strong as the ones online, and or will run into the regency bias. For example, it's very unlikely you'll actually meet someone that has used 10 different coffee makers…

>And I think it's this. The opinions you get from people 'IRL' are not apt to be as strong as the ones online, and or will run into the regency bias. No, I've noticed what he's talking about, and it's not the strength of the opinion, it's what the opinion is. Reddit has a moral system that's completely misaligned with real world morality, where having a child or being autistic makes you a bad person, even if you didn…

>having a child

Not sure if you've noticed but younger generations that go out and do things are big into having children. Now, some groups on reddit are more extreme on that, but the general trend of Americans at least is to tell the act of having kids to screw off.

---

But to answer your question, it's not Reddit that has this moral system, it's social media in general that has a moral system that does not match reality (kinda). The loudest idiots tend to get voted up, moderates disappear in the bulk of posts. Binary voting systems on sites tend to amplify this. Content suggestion systems tend to lift up contentious posts for engagement. Welcome to the internet.

But coming back to (kinda). This is becoming reality. Behavior IRL affects behavior online. Online behavior affects IRL behavior. People don't talk to their neighbors these days in most places. Communities are spread all over the earth.

Re: Reddit is OpenAI’s moat

#302

Post by rjh29 »

moffkalast wrote: post #253
rjh29 wrote: post #76

Earlier quoted context omitted.

Yes. For anything from opinions of movies/games to DIY advice to nerdy stuff like the best watch to buy in a given price range or the best synthesiser. Or people's opinions on a particular episode of TV! Or basically any hobby. It's not great for business stuff because there are way more people using e.g. AWS for their hobby project than running multi-million $ businesses on it. What are the alternatives? Facebook an…

> Quora is a shithole It's actually funny how much this is true. They used to be a quality source of info, but these days the default option is to show "all related" so you don't even get direct answers to the question you searched for, ads from top to bottom (and even somewhere in between) and just hordes of bots answering nonsense. At least they made it official with Chatgpt integration now lol.

Quora is a shithole because you can earn money answering, so people who have no knowledge at all are putting together convincing articles (probably with ChatGPT). And the mixing of unrelated questions into the answer stream is extremely annoying. I added it to my uBlacklist, anyway...

Re: Reddit is OpenAI’s moat

#303

Post by gingerrr »

importantbrian wrote: post #255
hospitalJail wrote: post #200

Earlier quoted context omitted.

>You'd usually be better off training on trade publications, scientific journals, or fandom than Reddit. Bold claim. There are plenty of engineers and doctors on reddit answering nuanced questions that arent talked about in any journals. That is what is missing. The absurdly specific stuff that reddit gets. Sure you can answer with nonsense that sounds realistic, but you could also answer with the exact text that sol…

> There are plenty of engineers and doctors on reddit answering nuanced questions that arent talked about in any journals. There are lots of really good subreddits that have expert feedback, but I think the issue is that going through all that and separating the wheat from the chaff is going to be a much more involved process than feeding in those other sources of expert opinion.

Ironic that reddit self-sabotaged the one popular place to converse with verified experts in their field so hard (AMAs) as that was probably the most trustworthy input source on the site considering participant vetting

Re: Reddit is OpenAI’s moat

#304

Post by dang »

93po wrote: post #300
dang wrote: post #252

Earlier quoted context omitted.

I wasn't the mod who added the question mark, but I'd guess it's because it makes a grand and speculative claim about the two most hyped topics of recent months. Nothing wrong with that, but in an HN context it generally makes for shallower and hotter discussion, and that is not the best fit for a site like HN which is trying to go deeper. (Note that I said trying, not succeeding.) The question mark is one device we…

I appreciate everything you and the mods do here. Enough so that I feel my past behavior on HN has been less than ideal from time to time and I am incentivized to improve it because I respect what you're trying to do here. However, the staff at HN are doing themselves a disservice and legitimizing criticism with this sort of action. If you feel this content is grand and speculative then downvote it or flag it like an…

I appreciate that you're working to improve your contributions to HN!

You might be underestimating how much curation/moderation we do of HN's front page. We do a lot. The front page is not generated by votes and flags alone (https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...). Upvotes and flags are important but the front page is a complex combination of those plus software action plus moderator action. If we didn't have the latter, HN would basically be impossible.

It's a pity—I'd love to manage without moderator action because the latter is tedious and it would be much more fun to write code to do it or just let the upvotes have their way.

Re: Reddit is OpenAI’s moat

#305

Post by HaZeust »

jedberg wrote: post #199
mbmjertan wrote: post #172

Earlier quoted context omitted.

Given that sama is a board member of Reddit Inc, and that this is happening after GPT-4 was trained on Reddit data, I wouldn't jump to conclude they're upset at OpenAI. SO had publicly available, no-auth-required data dumps. This makes it difficult for them to know who is using their data. However, this surely isn't the case for Reddit who offered only API endpoints for this content, and I'm guessing you couldn't use…

> and given that Tencent is a major shareholder in Reddit Snoop Dogg owns more of reddit than Tencent. The Tencent investment is very overblown.

Snoop Dogg's investment was $50,000,000.

Re: Reddit is OpenAI’s moat

#306

Post by jedberg »

HaZeust wrote: post #305
jedberg wrote: post #199

Earlier quoted context omitted.

> and given that Tencent is a major shareholder in Reddit Snoop Dogg owns more of reddit than Tencent. The Tencent investment is very overblown.

Snoop Dogg's investment was $50,000,000.

Indeed. And Tencent was at $150,000,000 three rounds later when the valuation was 7x.

Basic math days Snoop owns twice as much.

Re: Reddit is OpenAI’s moat

#307

Post by quickthrower2 »

WORMS_EAT_WORMS wrote: post #148
eggbrain wrote: post #42

I think from Reddit's perspective, they are extremely upset with OpenAI, in the same way that I'm sure StackOverflow is upset -- OpenAI took: - The entire corpus of data the community had curated over the last XX years - The "goodwill" that these platforms had developed towards third party developers in allowing developers to work with their data - Potentially large amounts of traffic that would normally come to thei…

Ah, maybe. I see it differently though. reddit has never been about authentic content. - When they started, I believe they boasted creating fake accounts to mimic engagement to help grow community - It is painfully obvious the amount of political, gov sponsored, and corporate astro turfing that “gets through”. Both human and bot comment farms are real and have no doubt been artificially bolstering ideas and content f…

There is a decent amount of good stuff on reddit. With a RLHF model trained on shitposting, political grandstanding, porn etc and use that to filter out, you are left with great stuff. r/experienceddevs. I would fine tune on that filtering out negative voted comments and you have a career advisor. Mods did the data cleaning already!

Re: Reddit is OpenAI’s moat

#308

Post by shp0ngle »

vultour wrote: post #151
shp0ngle wrote: post #110

Why does everyone says “moat” these days? I haven’t heard this term so over-used even 2 months ago

I have literally never heard this word before the infamous Google post, since then it's everywhere.

What google post? I think I missed that one

Re: Reddit is OpenAI’s moat

#309

Post by executesorder66 »

mbmjertan wrote: post #172
eggbrain wrote: post #42

I think from Reddit's perspective, they are extremely upset with OpenAI, in the same way that I'm sure StackOverflow is upset -- OpenAI took: - The entire corpus of data the community had curated over the last XX years - The "goodwill" that these platforms had developed towards third party developers in allowing developers to work with their data - Potentially large amounts of traffic that would normally come to thei…

Given that sama is a board member of Reddit Inc, and that this is happening after GPT-4 was trained on Reddit data, I wouldn't jump to conclude they're upset at OpenAI. SO had publicly available, no-auth-required data dumps. This makes it difficult for them to know who is using their data. However, this surely isn't the case for Reddit who offered only API endpoints for this content, and I'm guessing you couldn't use…

> I think this is more important to Reddit than ad revenue, as they could've simply built an SDK for probably less than this PR nightmare will cost us.

Who is "us"?

Re: Reddit is OpenAI’s moat

#310

Post by asylteltine »

EMIRELADERO wrote: post #79
ltadeut wrote: post #70

Earlier quoted context omitted.

> , that data can and will get scraped if the intention is to use for a model. How would that work from a legal perspective, though? Let's say there's no paywall and Reddit's terms of use disallow unauthorized commercial use of their data. Wouldn't that be a violation of Reddit's terms and liable to some legal procedure?

It is OpenAI's and Microsoft's position that obtaining, using for training, and displaying data via AI is fair use.

I really really really hope the copilot and other suits are successful. The idea that you can literally steal content in the name of “””AI””” and profit off it is just insane! How is it not like copyright infringement? The Warhol case is just one step behind training data. It’s basically the same idea.
Post reply on HN