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

Dead Internet Theory

kudmitry.com

Post reply on HN
Print view View original

681–690 of 735 posts

Re: Dead Internet Theory

#681

Post by oh_fiddlesticks »

leshenka wrote: post #535

I just hope someday we will have actual cameras adding provenance data to the image file, digitally signing it and its miniature version. Image editing software will then collect that data from all images used to create an artifact and append their miniatures and signatures to it. So as long this data is not removed you can verify that the image came from an actual camera and evaluate the degree to which it was edite…

https://roc.camera/

This was on hn this year, and it was, in classic HN fashion, dismissed as a problem in search of a solution. Well, perhaps people in this thread will think differently

Re: Dead Internet Theory

#682

Post by jibal »

thinkingemote wrote: post #381
jibal wrote: post #145

Such posts are identifiable and rare, disproving Dead Internet Theory (for now).

for now. Even this submission is out of date as images no longer have the mangled hand issues. We are actually blessed right now in that it's easy to spot AI posts. In 6 months or so, things will be much harder. We are cooked.

AI being harder to spot still won't make dead internet crackpottery true. As for us being cooked ... in so many ways, including literally due to AGW, exacerbated by LLM compute and by the orange demento's policies.

Re: Dead Internet Theory

#683

Post by agumonkey »

armcat wrote: post #563

What's changed significantly is the "intent" behind the content. The intent today is mostly the actual "content generation", i.e. "I write a blog post or a public github repo for the sake of having a digital footprint". This is vastly different when I used the internet in mid/late 90s. The intent back then was more like an "online journal" or something you hope someone else will find useful one day. Back then we had…

I remember when internet started to go from wild jungle to organized in structure and form.. At first I was delighted.. and then I saw how there was an industrial feel to it. Things were predictable (tooling influences you, google search algorithm too).. and whenever I found an old 90s website, you'd feel the time, craft, attention to detail, personal interest. pages and pages of long and interesting paragraphs. so crude yet so deep.

Re: Dead Internet Theory

#684

Post by 6thbit »

oh_fiddlesticks wrote: post #681
leshenka wrote: post #535

I just hope someday we will have actual cameras adding provenance data to the image file, digitally signing it and its miniature version. Image editing software will then collect that data from all images used to create an artifact and append their miniatures and signatures to it. So as long this data is not removed you can verify that the image came from an actual camera and evaluate the degree to which it was edite…

https://roc.camera/ This was on hn this year, and it was, in classic HN fashion, dismissed as a problem in search of a solution. Well, perhaps people in this thread will think differently

> verifiably real moments

would someone benefit from demonstraing a photo is real?

The top usecase I can think of it to ensure AI is trained on real photos. Any upside for humans?

Re: Dead Internet Theory

#685

Post by notatoad »

armcat wrote: post #563

What's changed significantly is the "intent" behind the content. The intent today is mostly the actual "content generation", i.e. "I write a blog post or a public github repo for the sake of having a digital footprint". This is vastly different when I used the internet in mid/late 90s. The intent back then was more like an "online journal" or something you hope someone else will find useful one day. Back then we had…

content generation isn't the intent, it's just the side effect. the real intent is to make money.

everybody is either trying to promote some product, or promote themselves as a "content creator" so they can start getting influencer marketing deals and payouts from youtube or instagram ad splits.

the internet was at it's best when most of the content on it was just because somebody had some information, and wanted to share it. that intent is still out there today, but unfortunately it's harder and harder to find buried amongst all the revenue generation.

Re: Dead Internet Theory

#686

Post by baubino »

spicyusername wrote: post #522
seiferteric wrote: post #116

My parents were tricked the other day by a fake youtube video of "racist cop" doing something bad and getting outraged by it. I watch part of the video and even though it felt off I couldn't immediately tell for sure if it was fake or not. Nevertheless I googled the names and details and found nothing but repostings of the video. Then I looked at the youtube channel info and there it said it uses AI for "some" of the…

divisiveness this kind of stuff will create I'm pretty sure we're already decades in to the world of "has created". Everyone I know has strong opinions on every little thing, based exclusively their emotional reactions and feed consumption. Basically no one has the requisite expertise commensurate with their conviction, but being informed is not required to be opinionated or exasperated. And who can blame them (us).…

> It is almost impossible to escape the constant barrage of takes and news headlines these days without being a total luddite.

It’s quite easy actually. Like the OP, I have no social media accounts other than HN (which he rightfully asserts isn’t social media but is the inheritor of the old school internet forum). I don’t see the mess everyone complains about because I choose to remove myself from it. At the same time, I still write code every day, I spend way too much time in front of a screen, and I manage to stay abreast of what’s new in tech and in the world in general.

Too many people conflate social media with technology more broadly and thus make the mistake of thinking that turning away from social media means becoming a luddite. You can escape the barrage of trolls and hottakes by turning off social media while still participating in the much smaller but saner tech landscape that remains.

Re: Dead Internet Theory

#687

Post by baubino »

gooob wrote: post #530
spicyusername wrote: post #522

Earlier quoted context omitted.

divisiveness this kind of stuff will create I'm pretty sure we're already decades in to the world of "has created". Everyone I know has strong opinions on every little thing, based exclusively their emotional reactions and feed consumption. Basically no one has the requisite expertise commensurate with their conviction, but being informed is not required to be opinionated or exasperated. And who can blame them (us).…

it's malware in the mind. it was happening before deep fakes was possible. news outlets and journalists have always had incentive to present extreme takes to get people angry, cause that sells. now we have tools that pretty much just accelerate and automate that process. it's interesting. it would be helpful to figure out how to prevent people (especially upcoming generations) from getting swept away by all this.

> news outlets and journalists have always had incentive to present extreme takes to get people angry, cause that sells.

As someone who’s read a newspaper daily for 30+ years, that is definitely not true. The news has always tried to capture your attention but doing so using anger and outrage, and using those exclusively, is a newer development. Newspapers and broadcast news used to use humor, suspense, and other things to provoke curiosity. When the news went online, it became focused on provoking anger and outrage. Even print edition headlines tend to be tamer than what’s in the online edition.

Re: Dead Internet Theory

#688

Post by pants2 »

dpkirchner wrote: post #656
pants2 wrote: post #106

Are there any social media sites where AI is effectively banned? I know it's not an easy problem but I haven't seen a site even try yet. There's a ton of things you can do to make it harder for bots, ie analyze image metadata, users' keyboard and mouse actions, etc.

It's impossible, so I figure we may as well accept it when browsing rando's content (or just don't read any of it) and reject it when interacting with friends and family. What else can you really do?

Sure, it's impossible to block everything, but you can at least make it difficult enough to post AI slop that the spammers go elsewhere. Most social media sites haven't even tried and are in fact actively encouraging AI slop.

Re: Dead Internet Theory

#690

Post by efreak »

femto wrote: post #155
therobots927 wrote: post #127

Earlier quoted context omitted.

Well then in that case, maybe we need a “vetted internet”. Like the opposite of the dark web, this would only index vetted websites, scanned for AI slop, and with optional parental controls, equipped with customized filters that leverage LLMs to classify content into unwanted categories. It would require a monthly subscription fee to maintain but would be a nonprofit model.

That's the original "Yahoo Directory", which was a manually curated page. https://en.wikipedia.org/wiki/Yahoo#Founding The original Yahoo doesn't exist (outside archive.org), but I'm guessing would be a keen person or two out there maintaining a replacement. It would probably be disappointing, as manually curated lists work best when the curator's interests are similar to your own. What you want might be Kagi Search…

The original dmoz/open directory project/Yahoo directory database is being hosted (but not maintained; I don't think it's being updated) at odp.org, which also includes links to other directory projects at http://odp.org/Computers/Internet/Searching/Directories
Post reply on HN