- cross-posted to:
- news@lemmy.world
- cross-posted to:
- news@lemmy.world
You must log in or register to comment.
Just wish we could destroy Israeli data centres… Russia is terrible but Israel is the new actually real Nazis
deleted by creator
You know what nobody in the news even mentions or talks about? The fact that multiple big AI “competitors” went offline at the same time a while back…
The empirical data shows a big decline in Russian bot activity overall. You see it on Twitter already. The strikes took down web crawlers and we saw that graph go down, but the spin of it was all just crawlers falls short of actually explaining a lot of what we’re seeing now. The ‘Amazon’ of Russia, probably didn’t keep any kind-of graph to show how their social media twitter bots are running public facing. Now the lie that it was all web crawlers has done it’s lap, and we’re all going to argue about this a little while longer.
I like how Musk’s Star Link cut off their service in retaliation over… Web crawlers?
Or diesel prices before the midterm elections.
How is it a lie? Do we have actual evidence that Russian bots went down?
We have only russia’s word on what was in that data center.
So you believe things are true, not when there’s evidence for it, but because Russia says otherwise?
To be fair, it’s too recent for us to have any robust data and analysis over it, but anecdotal and loosely backed claims do suggest a moderately reduced activity overall on contrarian accounts on popular platforms. The minority populists, depending on the country whether they are far right or some left wing niche.
I don’t believe the actual computational shape of a yandex data center would be right or efficient for bot farms, but i do think any external bot farms would use yandex’s indices and the currently trending hype word, “crawlers”, and all that, actual data center shaped work and resource, to carry out their work. The impact, if it’s real and not just imagined which is also entirely possible, would be the lack of resources for the bot farms to find the content to interact with and be efficient with the opinion shaping. That wouldn’t outright stop their work, just make it less efficient and convenient, which would make it slower and more laborious, which, if the anecdotes and such are right, would indeed show up as slightly decreased activity, not a full stop for the activity.
But right now it all depends on the anecdotes or the still fresh and dubious data claims, so it might just be there’s no real impact anywhere. But if there was an impact indeed, this would explain it. Nobody runs any sort of content mills manually, it’s all automated over something. Yandex indices seem entirely plausible as one part of that, and perhaps rented shared computational capacity to some extent too. And whatever else they might be using.
To be fair, I feel like either way it’s all blown out of proportions. The impact is definitely real, there’s real loss of compute and capacity, but I can’t see the bot farms and content mills being a primary symptom or anywhere close. Smells like populist and opportunistic media just doing their thing under capitalism and also the very same concept of opinion shaping on the “western” side of the conflict to me.
It is all anecdotes that can be explained by just a placebo effect. Actual bot farms (the kind of bots we’re talking about) don’t exist on search engine data centres. They’re set up in poor countries running on multiple second hand devices using proxies. Bots do exist, but you don’t get rid of them by blowing up a data centres used by a privately owned search engine.
That’s all normally the case, but we already saw evidence from leaked documents that people in the Russian government have paid for misinformation campaigns to influence elections, out of state funds. That’s a pretty significant ethical line to cross.
Running a sanctioned bot farm out of a data center that shouldn’t have one - wouldn’t be that much of a stretch, from what we already have evidence has been happening in recent years.
But in principle, I agree with you. It’s unlikely.
Not even just leaked documents. The geriatric cyberpedo released the proof himself.
This entire experience has soured me a bit on Lemmy. There were a flurry of excited posts about the massive drop that totally happened. Every article I read refrained from presenting any evidence, and all used waffling text that sounded like it was all coming from other sources that didn’t have any proof. It didn’t help that Zalenskyy seemed to explicitly say that there was a massive drop - and that sucks, because I tend to generally believe what Ukraine says. I know they’re in wartime and every government lies to some degree, but they generally seemed to be truthful for the most part.
So when I tried pointing this out in places, I got massive pushback - “No no, there’s totally been a massive drop!” Where? Where’s the drop? Where’s the numbers from someone doing actual measurements? No answers, just downvotes.
I see this time and time again - often where it’s probably less important, like people misread things and decide it means one thing when it doesn’t mean that thing - and usually, the thing is bad, so the fact that they completely misread it is probably not so bad, except that it means nobody is caring about the truth.
For example, the Vance couch thing bothers me. There’s plenty to string fascists up from trees for, but that couch thing was a joke in a book. So of all the legitimate things we can hate him for and joke about, that is a lie - a lie when it was taken as truth; originally a joke that became a lie when it was taken as truth.
We really don’t need to come up with bullshit to be angry. All it does is makes us susceptible to more lies. How much of what I believe is a lie? How much is just flatly incorrect because people keep repeating it as true and I’ve also fallen for it? I’m not immune.
It’s a serious problem everywhere, but Lemmy is small enough and generally progressive enough that it stands out because I EXPECT that fascists to lie. I would expect better from us. (And we aren’t as bad, but the amount of misinformation that is vibe-upvoted is insane)
edit: The only thing that makes me happy is that this article is getting upvotes, so there is some hope, at least.
You really gotta remember the ole internet nihilistic catchphrase… nothing ever happens. Even in traditional media, the “feel good” stories were often fake or misrepresented. People really want them to be true and will definitely latch onto the narratives.
Subjectively, I commented on how calm and pleasant a couple of my social medias had been before reading the news about 14 hours after the strike.
Literally everyone is susceptible to propaganda. Mostly all the old philosophers agree on one thing, you should be able to put down, entirely, everything you believe, so as to test you haven’t fallen for propaganda. Even those dudes fell for propaganda and realised they needed to continually self check.
THANK YOU. I have been saying this too. Thankfully those comments weren’t immediately obliterated so there are more sensible people still. But to your point, just because you want to believe something that doesn’t mean it’s real.
Lemmy is, on the whole, a good bit dumber than reddit
Reddit fell for the exact same thing.
Eh… In general, I can’t agree. And on this specific topic, I can’t agree either - I saw this on reddit just as much.
In fact, one of the reasons I hate it even more on Lemmy is that the general level of discourse is somewhat better here. Mostly. :P lol
I find that hard to believe since reddit is 99% bots and bots lack any intelligence
Welcome to humanity. The minority of “misinformation” we come across on the internet isn’t malicious, it comes from misunderstandings, jumping to conclusions, and hearsay. It’s existed before social media. I surf a lot of the “old web” and my latest find is a website that was telling me that a mummy was on the titanic, and it led to the sinking.
tl;dr the “bots” were web crawlers, not social media accounts. And there’s no evidence that disinformation campaigns saw any change at all.
Why did Musk cut Ukraine off from star link over web crawlers?
I think he cut them off because Donny wanted it over Ukraine not wanting to be in on his super cool “deal”.
Timing is sus
Did he? Seems unlikely.
No, there is no evidence and the graphs were just Yandex search crawlers. Yet there are numerous anecdotes about far right accounts getting substantially less amplification and engagement in the aftermath. That could be just a mirage but it could be more than that. That would warrant a proper statistical investigation at some point. After all, the question if Yandex servers host parts of the Russian psyop system is a valid one, that won’t be clarified by either misleading graphs, anecdotal stories or stories about how it was absolutely nothing, without offering any evidence for that either. Those big data centres were, from what I have read hardly just doing web crawler stuff.
Yeah I mean just because there’s no evidence doesn’t mean it’s not true, of course. But also there are potentially psychological explanations for this.
And notably, most of the actual news articles, even from the start, admitted that there wasn’t any proof if you actually read the articles. They were all “People are saying there’s a drop!” but farther down it was all “There’s no evidence it’s happening, but people are saying!!!”
One European Parliament MP said something, everyone else just went with it, Zelensky then doubled down and apparently that was good enough. It soured me on some publications for sure.
We were shown a graph and given a story, you’ve internalized it and spread it to others. I don’t that narrative because the evidence doesn’t play it out.
I did not spread it, I did not just believe stuff I still don’t see a clear picture of what was effected and if or if not online campaigns of the Russian state were in any way effected or not. They key for answering that would be in depth investigation of engagement and traffic on various autocratic right accounts in the EU, of AfD and Co. We know the timeline after all. Indirect effects are possible of course as well. Existing psyop campaigns can be theoretically also effected by knock on effects due to disruptions in the core online services of the country.
Very possible. They rely on good translations for example, and I would assume Yandex was one of the services that offered this. Or at least it might have run in the same data centers.
Without trustworthy translations, you can hardly influence as easily.
And even if there are alternatives, the existing data pipelines need to be reworked first (new APIs, no contracts, etc)
I haven’t done anything except explain the words in the article. If you have evidence, please feel free to share with the class.
And then Musk cut Ukraine off from Starlink
That’s not evidence, that’s just another article
I don’t know. I work at a cyber security company and the traffic to our site dropped by 58% overnight.
We were shown a graph and given a story, people are really gullible. It’s a shame to see a big win completely misunderstood by laypeople.
Is this the year 2000, we still measuring website traffic?
Hmm, every company monitors their web traffic. That’s standard operating procedure.
It’s just not a useful point of comparison anymore for the purposes it used to be. Does cloudflare track how many million hits I’m getting a day? Yes, but it’s next to useless as a measure for popularity or anything of the sort. You can’t differentiate human and bot traffic with any real degree of reliability, nor is there much meaningful reason to do so anymore with agents acting on behalf of real users. A year ago? Sure, I’d listen to an argument, but now users are bots too. Seeing “bot traffic” or web traffic move is meaningless now. A website with 1000 hits a day could be far more popular than a website with millions a day
It’s my job to do so
What a silly, silly comment
What do you even mean?
We’re also still measuring rainfall. Data doesn’t become less valuable as technology progresses…
Visits to a cyber security firm’s website couldn’t be less useful as a metric here
Yeah, and how exactly do you know that was social media troll bots and not like AI training spiders? Do you handle cybersecurity for twitter?
No, I manage the website for a cyber security company.
So when the Ukraine war first started, the executive team asked us to block all traffic from Russia to our site. We logged a similar drop in traffic.
This time around, the traffic also dropped but it looked like it came mostly from the UK, DE, DL and SG. It basically went to zero.
We cross reference those IPs and they were all from a VPN.
So you don’t manage a site with any social media component? I don’t think the bots that were crawling your site are the same ones people are saying disappeared from sites like twitter.
It’s not very surprising that banning all Russian traffic resulted in most of that traffic moving to a VPN; that’s the primary method of circumventing geofencing.
Well, I obviously can’t confirm if it’s related, but the day the strikes happened is the day my traffic dropped. It could be a coincidence. A data center could have both the twitter trolls and malicious traffic.
Yeah I suspect the traffic did drop when Yandex’s spiders and all the colocated AI systems stopped crawling your site; data from cloudflare confirms this. It’s not a coincidence; it’s just not what people are claiming it is. You’d probably see a similar drop if someone bombed data centers in the US, but that doesn’t mean they’re running social media troll bots. AI bots are responsible for a lot of traffic these days.
Do you have any information on what user agent strings the suspicious traffic was using? Yandex bots might still use their user agent string even if they were crawling through a VPN. AI bots would probably masquerade as a standard browser, but would be likely try to crawl your entire site unlike natural traffic visiting a few pages. With VPNs in play it’s too difficult to say for sure, though.
Sorry that you are getting downvoted you are exactly right. People can’t handle being wrong.
It’s pretty fair to assume any ml account is gonna push Russian propaganda, I’m guessing that’s the reason for the downvotes.
This story, that Cloudflare reported lower identified automated bot and API traffic from the ASN of the destroyed data centres, has been so massively warped it’s insane.
It demonstrates that average people and apparently tech reporters like toms hardware don’t know how the internet works at all.
When residential proxies affect at least a third of US households, you don’t need to route through a data centre that will be flagged as bot traffic. You set up a wire guard config and route to a residential proxy installed on millions of smart TVs, android boxes, smart appliances, desktop computers, and phones.
Seriously, I recently mapped multiple residential proxy networks with over 30 million unique US ips that were just scraping. You can use math to reliably fingerprint these networks, they’re big, with hundreds of millions of IPs involved, and you can rotate them via an API call to clean ones.
Zero of this traffic would be taken down by these strikes. We didn’t even see any perceptible change in internet traffic overall with this.
tech reporters like toms hardware don’t know how the internet works at all.
I’m sure they just don’t care. The vast amount of modern “journalism” is just reposting someone else’s story. And if they get clicks, they don’t much particularly care if it’s accurate.
Modern journalism, LOL. I’ve been reading on implementing Forths recently, there’s plenty of texts that don’t make sense, because how the hell do you change basic dictionary if you don’t have a dictionary already? There might be an implementation that doesn’t explicitly list that basic dictionary in the order of used ones, but what was literally written was wrong.
Even on things supposedly interesting only for people looking for specifics there’s plenty of vague and wrong texts. You start trying to approach the task by following them literally, then realize that’s not possible. Again and again. Until you just accept that you’ll get there faster if you make things up and do what you want, then add things you’ve missed along the way.
I’ve blocked my cameras and TV from reaching the wider web.
Is there a tool I can use to monitor my home network for this sort of thing?
Wireshark will give you deep infomation about any network traffic.
Hmm. That seems to be the big gun. And something that will require analysis.
I was hoping for something intended for just this purpose.
I would be looking at what tools opnsense provides. opnsense is a firewall operating system, its most effective to put it between your appliances and the internet. but some if its tools could also be used standalone
A proxy only routes traffic while the cpu cycles are happening elsewhere, allegedly, the data center that went offline.
Your argument only makes sense if those proxies are more than just routing network packets but also hosting the actual code of those bots. Oh, and you’d also need to assume there’s no c&c, the bots should be autonomous. Else there’s the single point of failure. If there’s a c&c then they can all go offline with a strike.
Exactly. Those smart TVs can be used as proxy to get around IP bans, but they can’t run LLMs on their own to generate social media posts. They can’t coordinate DoS attacks, only play a part in them.
I was arguing this point during the initial wave of suppositions by people just repeating vibes.
I feel like this really shows a certain flavor of social media bullshit very clearly.
The second problem, however, was that the ultra-specific claims about less activity in certain areas—MAGA, pro-Kremlin bot accounts, and AfD—had absolutely no basis in fact.
“We see no “dramatic drop” in traffic of Kremlin-aligned human trolls-for-hire targeting Russian domestic users (which we have been tracking for years),” wrote Antibot4Navalny, a group of anonymous researchers who track and respond to Russian disinformation.
Emphasis added.
“The party told you to reject the evidence of your eyes and ears. It was their final, most essential command.” — George Orwell.
Can you please provide evidence that shows that AfD engagement and traffic was not out of the ordinary during the start of the blackout? I am not saying that this was or wasn’t the case but I haven’t seen any evidence yet either way. So which evidence am I supposed to reject for the party?
Were you not on any of the platforms they target? Youtube alone became drastically less problematic in content and comments. That’s the evidence they want you to deny, it happened and it was really obviously different.
My Youtube is heavily sanitised and I avoid the horrendous search function. It barely finds what you are looking for anyway and is full of minimal effort slop and propaganda. If the latter decreased I could not say. I am not a masochist.
Mine is as well, but that’s never stopped youtube from trying to push far right/manosphere/conspiracy theory garbage at me I then have to block and remove from my history. Until the Russian data centers got blown up when it stopped like someone turned off a faucet.
first comments I saw in the article was people here saying. did we see a reduction? I think because it did not feel like it especially with the increase with the election coming soon.
