Yeah, I never know if people write genuine, stereotypical posts like that. Or if that’s bot activity. Being adamant not to be a bot is kinda what bots do.

Haha, yeah, pick your poison. Though we’re a diverse crowd. I really appreciate how there isn’t an endless stream of doomscrolling for me. My feed just ends at some point and reminds me to resume with other things 😅 (Also less noise, US politics and mainstream memes, which would be the real reason why I’d advise not to scroll All.)

Seems to me like the big tech companies are way ahead anyway. Google probably has enough data collected to know whether someone (or something) is eligible to watch a video. At least they distinguish based on network, maybe behaviour and in doubt they’ll force me to log in and present my session cookie. That’s all bad for privacy as is.

I wouldn’t be surprised if they follow up revamping the captchas. Maybe in the future it won’t make you click on pictures with an omnibus in it, but directly talk to the Chrome browser on your device and make that (or some hardware chip in the computer/phone like a TPM) verify you’re a human. Of course that one might as well contain a unique ID and an advertiser ID… I think I read on the news how they’re considering to redo the captchas. Not sure if it’s going to end up like that. But my best bet is it’s gonna be bad in some way or another. Perfect opportunity to tie something down for the users, remove use-cases that aren’t consume only, and track them some more.

From what I read on the news, they in fact caught some people by looking at other Flock cameras. Be careful anyway, while doing illegal things. And don’t brag about it or tell your friends.

Yeah, unfortunately the Fediverse isn’t as big as I’d like it to be. I can’t find people to talk about some of my interests, either. I’ll occasionally put in keywords on https://lemmyverse.net/communities , have a look at PieFed’s curated feeds and topics (Edit: another topic list)… But there’s probably just not enough people here who are into miniatures…

This creates perverse incentives for sycophancy to persist: The very feature that causes harm also drives engagement.

I bet that’s also why AI is like that (sycophantic). I don’t see any technical reason why a language predictor needs to write overly agreeable text. Or be as bland and repetitive as they are. That’s probably because they’re designed (by big tech) to be like that. Due to user preference.

Idk, I used to be fascinated by language models early on, for example when Meta’s first LLaMA model weights got leaked. And the time after when we got new models and discoveries every other day. And if I remember correctly, they weren’t all like that?! I distinctively remember a few language models which were more or less agreeable than others. Some would respond to a question in a single paragraph. While ChatGPT loves to write pages of redundant stuff and claim to highlight every side of each coin.

But with each new iteration, all the language models homed in more and more on the “helpful assistant” style. Possibly pioneered by ChatGPT?! At least that one felt from the get go like it was supposed to treat me as a 5yo who needs a lot of fake positive reinforcement.

Delete the All feed! It’s just stupid. Subscribe to what you like. Maybe have a short glimpse at the All feed on a biweekly basis for 5min, but that’s it.

I feel at this point it doesn’t even matter. I’ve recently renovated parts of my IT infrastructure… And I mean they’re not only harming news outlets and Wikipedia… Everyone is under attack. They’re scraping my sites, searching for all kinds of stuff until the database process catches on fire. They put weird stuff into every web form, send plenty spam mails… Sign up random people for your newsletter, unsubscribe people from your newsletter… The OpenClaws are able to solve the CAPTCHAS and/or sign up… It’s crazy. Law aside, I don’t think I have any option but to block them myself. This will also block legitimate crawlers because I have to put in place some drastic measures with some services. But at least I can continue to run my website.

I see other people do the same. And I also run into rate limiting all the time. Back in the day, my Home Assistant could just query trashday, whether the train is on time, which flights are overhead… I could just run yt-dlp to save some YouTube videos. That’s all becoming more and more difficult because of all the countermeasures. There’s still some chance to force people to sign up and offer some protected API…

But whether someone passes a law or doesn’t… I think the internet admins are still going to (have to) kill the bots.

Guess they’re mostly not paying up, then. That’s just deducting the difference from the rest of the tip jar until there’s nothing left from a given day. And only then are they forced to chip in?!

Hope it was a Golf 2 Fire & Ice edition…

Fascinating. Do people also write down the amount of tips received each hour? To make sure the boss pays for the slow hour, and it doesn’t end up coming out of your own pocket anyways, from the other more busy hour? I can’t believe Americans are doing that amount of maths?

When I was a pizza delivery boy, I just did the cashing-up once a day. The cash register would know how much of the money in my pocket was owed to the boss, and I got to keep the change. Only time maths was involved was when the delivery cars broke down (which they did all the time) and I had to use my own car and the boss had to compensate me for using my own car.

Most outrageous thing to me as an non-American is, how his boss is allowed to pay him $3 an hour for doing the job.

Dunno if that’s “on purpose”. I think it’s more, back in the day .ml was the defacto standard instance, so lots of average people went there and you’d find a wider bandwidth of people. That might have changed with some people leaving and the new visitors not signing up on .ml anymore. That makes the rest of the group condense to people who pick .ml on purpose. But that’s not a deliberate plan, more how maths works. 😅

Yeah dunno. Did they follow instance rules? It states inflammatory posts need a Content Warning and I can’t tell whether the author is in violation of the instance CoC or following the rules?! Seems there is no such warning on the screenshots. We could probably ask the admin or mod in question. Maybe it was a mistake, maybe it wasn’t… We don’t really know any details besides how Kazim thinks they’ve being wronged.

Thx. That article is way better. And they even have actual links to the sources…

And what exactly is is the real number according to Cloudflare? They’re hinting at it all the time but never telling the actual number?! Is it 9%, 2%? What kind of article is this? Isn’t the one job of journalism to tell us what reality is? But they deliberately(?) omit this one thing?

I wish my friends and family did the pre-screening of messages and their importance to me as well 😆 I get important messages and silly memes all mixed in the same group chats, so it’s impossible for me to tell whether I need to pick up my phone.

Background will do. I think even swiping it away is fine, as long as you don’t force-close it. Just pay attention there is some Play store or Aurora store doing automatic app updates. I skipped on the updates and once WhatsApp hadn’t been updated for several months, they closed the connection.

I run continuwuity and the mautrix-whatsapp bridge. It’s not 100% reliable, but close enough for me. Especially after I moved it from the NAS in my basement to a VPS in some datacenter. I occasionally had some issues with my ISP doing maintenance, a power outage at home… I let WhatsApp run on my kitchen tablet 24/7 so it keeps that 14 day timer running. To my knowledge that’s unavoidable unless you switch WhatsApp Business or something.

And I like Continuwuity way better than Synapse and the ridiculous amount of resources it uses to forward text messages.

But yeah, silently dropping the connection isn’t great at all. I think the only proper solution is to nag people into using a different messenger. Signal is relatively straightforward to use, made for average people and comes with the same functionality as WhatsApp.

Thanks but I guess regular Stable Diffusion is a bit too old for my taste. Seems we have better, more efficient models as of today and they’re way easier to prompt, can do text, the right amount of fingers… And all of that requires constant fiddling with the earlier models.