Hacker News Reader

Comment by rchaud | Hacker News Reader

It's over.

It isn't just Twitter, it's every single website that's turned themselves into a login-walled "application".

Twitter's relative openness lasted a long time. It was open by default because it is a product built in 2006, when the idea of coralling people into walled gardens to show them ads didn't exist.

Apps built later take the concept of "walled garden" as a default feature. Slack , Discord, Snapchat, Tiktok, Telegram .... all largely closed off platforms. You can't see anything unless you're logged in.

View on HN

But Twitter is an ad-driven platform. Most of those other ones are not. It's ALL about the exposure to eyeballs. But now that Elon chewed on some wires and killed whole colocations, they are not able to serve all that traffic that is supposed to view the ads - for all the talk about how "Twitter is still running like it used to" after being gutted like a fish. It is definitely not.

It's not really that ad driven anymore since most of the advertisers have gone to fuck themselves as musk told them to :)

But seriously, if they want to increase revenue it probably is more effective to not insult their advertisers than it is to make a login mandatory.

Once upon a time, you could go to Twitter as a non-registered user by clicking on an embed in another application, view the post, and then click on the user's post and see all their interactions without ever registering or logging in.

Today, you get to see the one post, and if you're lucky, it will give you a few of their most highly interacted posts. If the point of twitter is to make money, then it doesn't matter whether someone is logged in or not: ads in front of eyeballs is what's important. Musk's entire schtick since buying the platform is that he'll bully people into doing what he wants. But users on the internet have near zero patience with such demands from platforms, especially when the platform in question is losing relevance and isn't unique. What genius! Not only did he make the ads less valuable, but his behavior has created less competition for the ad slots that people actually see.

I've watched while shaking my head as a former employer has taken a similar approach and absolutely cratered their online business. I warned them that it was going to end in disaster. I provided the evidence that we'd used to make decisions against exactly what they were attempting to do. I can't unstupid them.

> If the point of twitter is to make money, then it doesn't matter whether someone is logged in or not: ads in front of eyeballs is what's important.

To serve you ads that have any hope of making money, they need to know which eyeballs they're going in front of.

No they don't. People sold ads profitably for hundreds of years without any sort of surveillance, and many people continue to do so today. If it were impossible to make money just selling ads (I wish it were) then billboards and bus stop posters wouldn't exist. You can even make money on the internet this way: www.ethicalads.io.

Twitter also has an enormous advantage: even if they don't know anything about you, they know a whole bunch about whoever posted the tweets you're reading. They could do a lot of targeting by just looking at the content you're viewing.

> People sold ads profitably for hundreds of years without any sort of surveillance, and many people continue to do so today. If it were impossible to make money just selling ads (I wish it were) then billboards and bus stop posters

Billboards and bus stop posters are cheaper to run than Twitter.com is, so making much less money is ok.

The whole reason Twitter has always been unprofitable but Facebook/Meta is a trillion-dollar public company is that FB figured out how profitably identify users and serve them ads and Twitter didn't.

I agree with you, it is unquestionably more profitable to spy on people and then sell them ads. My point is that you can still make money without the spying part, and there is no reason for us to accept that the two always have to be bundled. If you can't make a profit without spying on people then your business model sucks. The fact that you might be able to make more money by being less ethical is not very interesting, that's true of almost anyone.

Network television is enormously expensive and entirely paid by ads without knowledge of the viewers. Instead they make assumptions based on surveys of demographics and manage to be very profitable.

I wonder if you could make a comparison of ratings of a TV show to active users of a website. For example: is Twitter the equivalent of daytime TV, or multiple series-finale-of-Friends per day.

*edit: it looks like they are the equivalent of 0.8 Friends finale / day. 41m daily active users vs 51m viewers of the Friends finale.

Cookies do this for them.

It absolutely matters if someone is logged in. 3rd party cookie blocking on Safari, Chrome and Firefox has crushed the ability for advertisers to do remarketing with any level of precision. That was where you didn't need the user to be logged in, you just needed to match their cookie ID on Twitter.com with the cookie ID passed on by Site X, which the user had visited previously.

Nowadays, Advertisers upload customer lists with email, phone numbers and other identifiers into Twitter Ads for ad targeting. Twitter shows those advertisers' ads to logged-in users that have the same email/phone # or other identifier.

https://business.twitter.com/en/help/campaign-setup/campaign...

Anything else is effectively untargeted and useless.

They already have the only information they really need, which is what you've requested to look at. If I'm following someone who tweets about X you can safely guess that I'm interested in X and target me with ads related to that without ever knowing my identity.

There may be a reason there's no social media network that does this.

To some extent they very often do use the type of content you consume for various things, but they don't use that data exclusively because they can make far more money by collecting, leveraging, and/or selling every scrap of personal data they can get their hands on and money is really all they care about.

It may be that it wasn't feasible on large scale to match content to ads. I feel like recent advanced in machine learning and ai could change the game there.

What were describing is essentially how television ads worked for decades, isn't it? You know the content, so you can identify the demographics and advertise properly. Podcast and YouTube sponsor deals often work that way too (not always, but that's on the advertiser).

Interesting that something initially seen as a privacy victory has ended up harming the experience of anonymous internet browsing.

Still a privacy victory, and if they want to annoy me I'll just go somewhere else, that's fine.

> Musk's entire schtick since buying the platform is that he'll bully people into doing what he wants. But users on the internet have near zero patience with such demands from platforms

The continuing decline of reddit makes me suspect people are much more tolerant of being bullied than you think. when old.reddit.com is finally killed off (it already seems to be getting less functional) I'll stop going there entirely too.

It’s somewhat intriguing how long they’ve kept old reddit alive. I had figured it’d get the axe fairly soon after they made the redesign the default, even if they claimed otherwise. I suspect there must be a contingent of users who are big contributors or otherwise important, who they believe will leave if they shut it down.

I suspect they have loads of data showing that contributors use old Reddit and cat video watchers use new Reddit.

And they haven't chosen to cater only to the watchers, yet.

People ceaselessly cry about reddit and its moderation while anyone is perfectly free to start a community of their own and moderate it to their own heart's content. I've seen no real decline in reddit. It's generally the same mindless shithole it was 10 years ago, while the communities which have had steady moderation haven't changed at all.

The quality of individual subreddits has a very wide range, but it's the admin side of things that just get worse and worse. Anyone can create a community but you can't escape the way they run the site.

Over time more and more subreddits get shutdown for questionable reasons, censorship creeps in more often, long standing problems with moderation go entirely unaddressed (both in terms of problematic moderators/moderation and in terms of the lack of tools/resources available to moderators), transparency, communication, and integrity get worse all the time, nice features get removed, obnoxious new ones get added, etc. Even old.reddit.com is starting to become broken.

If you're lucky you can still find some great little subreddits, and maybe they'll even be allowed to stick around, too many good users won't get chased off, and they don't degrade in quality much as the site itself gets worse around them. There's still no good reddit alternatives but I'm still hoping we'll get one soon.

Well it probably doesn't help that they're now riddled with bots. About 6 months after the Musk takeover I've started to receive 2-3 likes a day from bots and a follow every week. I only have an academic account so my bot interaction vastly outweighs my non-bot interaction. I stopped reporting them because they still exist.

The big change I saw was that they no longer suggest to me 30 other bots with the same profile picture when I click on the bot's page but instead suggest I follow other researchers. I guess a slight improvement, but only visually. It's weird to see him talk about how bots are so harmful to the platform, see that the platform is identifying bots, but then not see any action. These are accounts that can be identified via a naive bayes filter too, so pretty low hanging fruit. But then again, same thing is happening to my gmail.

There is a financial incentive to ignore bots.

The investment banks and advertiser's want to know how many eyeballs the site has, and counting bots too increases that number and hence the company valuation.

I'm more taking about the hypocrisy since Elon made a huge deal about bots

Slack, Discord, Snapchat, Telegram are not open forums as their main selling point. They are 1 to 1 or group/community chat applications first. So relatively scoped or even private. It's like a blog vs icq/irc.

And yet Discord has become a de-facto forum replacement for niche topics/communities.

...Which is awful.

As is often said here, Discord is where information goes to die.

It's horrible that a lot of web development tools and libraries keep the discussions only on discord.

Add Reddit to the list. Lately, I get "Whoa pardner" treatment whenever I try to read it without logging in. But at least you can still create an account without a phone and without even an email. AI is evil and already makes us all suffer.

Forget them all. If it isn't on the open web it's a private club.

Actually in case of Telegram you could open in browser public channels.

Example: https://t.me/s/durov/ https://t.me/s/telegram/

It’s because of AI scraping.

Wait until you realize ‘verified’ paid accounts will be a necessity in the future to prove personhood unless we toss it all in the wind and create a national id system which guess what, you’ll be tracked by.

Which way brave one? You will lose an arm or a leg. You get to choose

> Which way brave one?

Probably the same way my grandparents did.

Why should Twitter care about AI scraping, beyond the load it causes to their servers? Scraping isn't new, and I don't think the overall rate of scraping has changed much since the advent of AI - it only needs to be scraped once, then it's in a dataset (I don't have any stats to back this up, though)

If you have a company building AI on top of Twitter's data, it could've been a company Twitter itself could've built. It's the feeling of lost money for them.

The API has been wide open until 2023, so if there was anything of value, I think we would have seen something by now. Musk would also have been focused on that instead of trying to become WeChat US.

According to Musk, below:

Almost every company doing AI, from startups to some of the biggest corporations on Earth, was scraping vast amounts of data,” Musk said. “It is rather galling to have to bring large numbers of servers online on an emergency basis just to facilitate some AI startup’s outrageous valuation. Several hundred organizations (maybe more) were scraping Twitter data extremely aggressively, to the point where it was affecting the real user experience."

To address extreme levels of data scraping & system manipulation, we’ve applied the following temporary limits:

- Verified accounts are limited to reading 6000 posts/day

- Unverified accounts to 600 posts/day

- New unverified accounts to 300/day

I think you're overestimating the value of raw text.

A large unsorted dataset is not "free" for AI companies. Without API access there's very little useful information that can be scraped from a tweet.

In my opinion, Twitter just wanted to convert lurkers to make it look like the platform isn't dying.

As if paying 5 bucks is going to be tough for an agent.

I'm not convinced AI scraping has anything to do with it. A company resourceful enough to develop & train AI is fully capable of setting up infrastructure to massively create bot accounts to bypass the loginwall.

You can see plain text Telegram (and Twitter?) posts by a direct link without login.

You cannot do that anymore to FB or Instagram posts.

twitter is the best shit that startups culture ever produced for developer healthyness

built with rails

full restful API and good docs

You must be new and young, most internet communities have always been login-walled or otherwise inaccessible to the wider public.

Why? Nowadays it's to combat freeloading "AI" data scraping, but more predominantly it's been to fight spam and scraping of data in general.

So no, nothing is over. If anything, we're getting back to the status quo of 20 years ago when most forums required you to make an account to enter.

Here's a list of open-access communities I grew up on:

Hacker News, Myspace, LiveJournal, XDA-Developers, RateyourMusic/AbsolutePunk forums, and about a thousand phpBB/vBuletin boards from 20 years ago. Every single one of these is close to 20 years old.

It is in fact the "new and young" that think they need to "download the app" and "join the community" for every single thing that could just be a Discourse message board.

Hell, there were BBSs that let you browse around under a guest account with zero requirement to register an account

It's as if you were describing an alternate reality

Extremely few forums required an account to read them 20 years ago, that practice largely began with Facebook

Most of the forums I patronized left their index open to the wider public, maybe also let the wider public read posts, and almost always blocked user profiles (read: treasure troves of personal information) without logging in.

User profiles were indeed blocked to the public on most forums, but I very rarely found that a problem

Change that maybe to definitely, I guarantee you that very few were auth-walled

> Extremely few forums required an account to read them 20 years ago

The few forums I currently frequent don't require an account to read most of the threads.

Reply on news.ycombinator.com