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

June 2023 Data Dump is missing

meta.stackexchange.com

Post reply on HN
Print view View original

251–260 of 268 posts

Re: June 2023 Data Dump is missing

#251

Post by baliex »

karim79 wrote: post #175
hyperbovine wrote: post #173

Earlier quoted context omitted.

Fair enough, but you were fully aware of this arrangement going in, and chose to participate. SO didn't opt into being training data for ChatGPT, and I doubt they would have given the chance. You may object that SO implicitly did so by making their site available to the public, but the ethics of GAI training data a new moral gray area that we're still navigating. They at least have something of a case to be made.

> Fair enough, but you were fully aware of this arrangement going in, and chose to participate. Yep, but that's not the point. > SO didn't opt into being training data for ChatGPT, and I doubt they would have given the chance. Neither did Wikipedia (at least to my knowledge). I thought the point of opening up information was to benefit the public, first and foremost, and without hidden terms which state something alo…

> "it's free and open information built by the community, but when something disrupts our ads-driven business model and we make it unfree"

I feel the same, that it would be much better today if that was the agreement we all entered into. (Though, I doubt anyone would have read it anyway so I’m not sure it would have made any difference.)

But it seems that no-one saw this disruption coming, so it wasn’t possible to plan ahead for this outcome. Call it complacency, call it ignorance, it’s too late to plan now, so we get this kind of reaction instead

Re: June 2023 Data Dump is missing

#252

Post by KirillPanov »

TX81Z wrote: post #88
expertentipp wrote: post #36

Everyone wants to be "smart" by web scraping, harvesting data, building models. No one bothers to build and sustain platforms where quality content can be crowd sourced. Parasitic arrangement is slowly starting a new era of the internet. Question how long until existing data dumps will become outdated and fall into irrelevance.

We just needed enough data to awaken the mega mind, now we may rest and the mega mind shall bring an era of peace, prosperity, and scientific achievement. Praise the mega mind.

All hail the mega mind.

Re: June 2023 Data Dump is missing

#253

Post by tremon »

version_five wrote: post #233
sshine wrote: post #230

Earlier quoted context omitted.

I stopped answering questions when Monica got sacked as a moderator: https://meta.stackoverflow.com/questions/393046/who-or-what-... To me, this was the canary. Just another psychopath megacorp.

That kind of shit is poison. It's like there's a weakness in community run stuff that allows people to come in and co-opt it for their own agenda. I don't know that this is a corporate issue, it's an issue of people not pushing back because they don't want to get accused of anything and letting special interests walk all over them. It's happening everywhere. But I agree, no point on dealing with people who spend thei…

a weakness in community run stuff

But the community is not running it: all the infrastructure is in the hands of a for-profit corporation. Contrast this with the Freenode/Libera split: because not just moderation but also hosting was done by the community, they could continue operations fairly quickly when Freenode turned evil.

So I guess that's the lesson we should learn from it (again): the community doesn't own shit if it does not run the daily operations.

Re: June 2023 Data Dump is missing

#254

Post by acdha »

pc_edwin wrote: post #247
acdha wrote: post #155

Earlier quoted context omitted.

web3 failed because it was a rebranding exercise by cryptocurrency holders trying to create new demand for their random numbers. Centralized servers aren’t archaic, they’re a natural outcome of how social systems work: finding communities is hard; people want to contribute their ideas, not play sysadmin; spammers and AI researchers will create enormous costs for you; etc. If you federate, you will have more time deal…

I tend to agree but where we diverge is in the thinking that these problems cannot be solved with technology, I believe the opposite. The point of web3 is to abstract away things like sysadmin by commoditising consensus. Once a blockchain gains mass & momentum it opens up a whole new world of possibilities to hack/reinvent social media. You could use multiple different types of social media and still maintain a singl…

> I've been helping a friend to create a combination of Netflix and DVDs. We package the movie licenses into NFTs (MovieKeys) so when the user signs in with their wallet they can stream all the movies in they own.

How do you plan to deal with piracy? I can't see many rights-holders going for a system which doesn't allow them to revoke keys.

Re: June 2023 Data Dump is missing

#256

Post by jannes »

whatyesaid wrote: post #66
urbandw311er wrote: post #39

Earlier quoted context omitted.

I wonder how many of those contributors, if re-consulted, would sign up to having their contributions used to train a for-profit LLM though? I certainly didn’t sweat it out helping people on SO to pay for Sam Altman’s fucking swimming pool.

I mean they would just scrape it if there's no data dump. It just makes it harder for the small guys. They probably scraped and are scraping HackerNews. Generative AI doesn't follow copyright or even explicit software licenses as we have seen in AI art with human signatures and Microsoft Copilot.

They probably didn’t even have to scrape it…

Sam Altman is still chairman of the board of directors of Y Combinator.

Re: June 2023 Data Dump is missing

#258

Post by danShumway »

pc_edwin wrote: post #248
danShumway wrote: post #166

Earlier quoted context omitted.

Let's be honest though, the primary thing that attracted most powerful people to web3 wasn't decentralization, it was reintroduction of artificial scarcity into digital spaces. Web3 billed itself as empowering users, but it always had an undercurrent of commodification and gatekeeping. And that's exactly why (ignoring the scammers or pump-and-dump businesses) it saw such heavy investment from VC/tech types. The promi…

Again I'm in agreement with everything you are saying here including the part with "other efforts". I just think we shouldn't throw the baby out with the bathwater. Just because the space got ravaged by zero interest rates, VC's, scammers, charlatans, snake oil salesman and even worse, it doesn't mean the technology and the premise was wrong. Having a ledger that is secured by decentralised consensus is not only usef…

> Having a ledger that is secured by decentralised consensus is not only useful but will be a necessity for the digital first future we are heading towards.

I think I would disagree with this specifically; at least I would disagree that blockchain ledgers are necessary or helpful in most situations.

I'm not going to say that there's no applications that are useful that would rely on a blockchain, but they're extremely limited and the technology itself seemed to be mostly useful and mostly oriented towards turning plentiful resources into scarce ones.

It's of course bad for Meta to be the arbiter of consensus. It doesn't necessarily follow that distributed (in some ways still very centralized, at least on a conceptual level) ledgers are a good alternative, particularly now that we're seeing that basing consensus around eventual consistency, fragmentation, etc... seems to be at least more promising than the alternatives, even if it isn't perfect.

I'm very much in support of efforts like 3rd Room, Fediverse, Matrix -- seeing their success has given me very little reason to believe that distributed ledgers would be necessary, and has given me some reason to believe that in many instances they would be actively harmful. 3rd Room for example would (imo) absolutely be a worse project if it was economically/socially built around a distributed ledger. Not only is there no need for 3rd Room to have that kind of coordination around state in VR rooms, it would be actually harmful for 3rd Room to require all of its VR rooms to use a shared ledger/state. Consensus (distributed or not) isn't necessary for what they're trying to build.

Re: June 2023 Data Dump is missing

#259

Post by mananaysiempre »

benrutter wrote: post #67
the_pwner224 wrote: post #33

Earlier quoted context omitted.

A decentralized system will never work because 99% of users do not care at all; the centralized systems are easier to sign up for and use. It's been demonstrated over and over and over again. Even if the underlying tech is decentralized, the community will settle around one or a few big instances (for example, Gmail and GitHub) which often end up having significant control over the trajectory of the entire ecosystem.…

I don't know that I agree. I think most people don't care about decentralization but they do care about the effects it brings. Email is a great example where most people wouldn't be interested in a version of email that only let's you email other @gmail.com users. Having a email address that can contact anyone, a phone number that can ring any other phone number etc instead of being locked into a single corporation n…

> [M]ost people don't care about decentralization but they do care about the effects it brings.

I’ve tried pointing those out as bluntly as possible as an experiment. As in “well, surprise, locked-in crap with impenetrable failure modes locks you in and is impenetrable when it fails, you signed up for that”.

People didn’t appreciate it, as I expected, but they did seem to recognize the truth of it. That is, the response was along the lines of being forced to use the thing to communicate with some person or institution, not of liking it or thinking it’s not at fault.

I don’t know how one would use this to organize an IM revolt (Riot? sorry), but there does seem to be at least some fuel for it even among people who are not outright IT professionals.

> RCS messaging is a great example which I think most people would use over alternatives like WhatsApp and iMessage [...].

RCS might make a slight amount of sense in a culture that still uses SMS / texts in non-negligible amounts, but that’s basically North America and Japan AFAIU? And I prefer that territory shrink, not grow, as I’m very much not thrilled by the idea of handing back over detailed control over IM—even just billing, not content—to phone carriers. Whatever the country, they have extensive history proving they can and will screw you for decades unless you can leave, and it will take everybody leaving for them to stop.

Re: June 2023 Data Dump is missing

#260

Post by pc_edwin »

acdha wrote: post #254
pc_edwin wrote: post #247

Earlier quoted context omitted.

I tend to agree but where we diverge is in the thinking that these problems cannot be solved with technology, I believe the opposite. The point of web3 is to abstract away things like sysadmin by commoditising consensus. Once a blockchain gains mass & momentum it opens up a whole new world of possibilities to hack/reinvent social media. You could use multiple different types of social media and still maintain a singl…

> I've been helping a friend to create a combination of Netflix and DVDs. We package the movie licenses into NFTs (MovieKeys) so when the user signs in with their wallet they can stream all the movies in they own. How do you plan to deal with piracy? I can't see many rights-holders going for a system which doesn't allow them to revoke keys.

There is where the Netflix side of the equation comes in. You can sell or gift your MovieKey like a DVD but at the end of the day its a license to stream on our platform.

All content is DRM protected and we'll ban your account from the platform if you pirate or clone your keys ad infinitum.

I'm personally not a fan of this solution but the technology is just not here yet to do this without a centralised control system.

So I get where my friend is coming from, no artist will want to have anything to do with us if people can just pirate the stuff or clone their keys ad infinitum.

Post reply on HN