« first day (4696 days earlier)      last day (551 days later) » 

00:39
Been gone for like.1 day or so. Is the strike still going on?
GcL
GcL
01:10
@M.A.R. Few things add urgency to a moderators want to talk situation like a steaming pile of spam and troll posts does. Plus, something like that might be a landmark that could persist in corporate memory. "Remember the last time we didn't bring our concerns to the community mods first? Steaming pile of troll posts."
@BigJoe yes
Negotiations ought to be starting soon
01:46
We have a new member on Astronomy.SE with an obscene user name. I'm almost tempted to temporarily break the strike & flag it. I think we still have 1 semi-active mod...
GcL
GcL
@PM2Ring Like Copernicanus? Galil-ahole? Smooth_black_hole? g_string_theory?
@PM2Ring Leave it. I'm sure the SE staff have things well in hand
 
1 hour later…
03:21
@JourneymanGeek Where is your diamond ? nvm i see it
@BigJoe >_>
 
1 hour later…
04:59
@Feeds The meeting that changed how we build community
 
2 hours later…
06:35
@ShadowWizardStrikesBack nothing much. Just that when Tink mentioned she's heard horrible things from people who are supposed to be spiritual icons for children, and I've heard quite a few things of my own. You give idiots a vulnerable audience (kids) and free rein and they'll . . . They'll say very interesting, disinhibited things
Things they probably wouldn't even say if intoxicated
@MetaAndrewT. there's that c-word again
@GcL I'm not convinced they're as concerned about whether or not piles of troll posts are aromatic. That's usually a curator/moderator thing. People high up there know though that if unpaid help doesn't mop the floor paid help has to, and that's probably what's determined/s the success of the strike
@BigJoe on Android it looks like a big red wart and hard to miss
@JourneymanGeek <_>
Don't laugh, I have an eye disorder
07:03
@M.A.R. Not really happening en masse here, but it's definitely... something.
07:37
@PM2Ring unless something really extreme e.g. cp then what @GcL said, no big deal. IMO
@GcL why always black?! So racist. :D :D :D
@BigJoe in his pocket. The small one of the jeans, ya know. ;)
@MetaAndrewT. and onebox on a strike.
@M.A.R. yeah, and I keep getting memes (?) showing hidden meaning, smartly hidden, in children books, or TV shows, and not only old from the times when nobody cared. ;/
(in silly Line groups, though most are crap, some are actually correct and you can actually see it's not just a theory.)
08:04
@ShadowWizardStrikesBack what are you talking about? I'm not talking about innuendo or whatever it is you're referring to
@Tinkeringbell well, to be fair, the idiots here are at least not too malicious or violent. It's just their super weird, super ugly conspiracy theories about Jews/Freemasons/their neighbor/KGB/whatever the heck people in Japan are doing right now/CIA
@M.A.R. Oh. Mine weren't as conspiratorious, just like 'you're supposed to love everybody. Except... '
'No gossiping. But please stay around after service to do exactly that'
@Tinkeringbell oh. Those come at high school. There's a whole procedure to it you know.
Very ineffective but still pretty procedural nonetheless
Oh, for sure.
The face on my mom when I give money to a homeless person... priceless. "But he'll just use it for drugs" "But you raised me to give if someone asked"
"Yeah but not like that!"
"Says it where?"
Junior high is about a Basiji principal saying Americans made jeans so they would constrict your balls and make them shrink.
@Tinkeringbell mine is way more understanding because she's a healthcare worker, but in practice she also, for example, sees addicts more as criminals than patients
Tabriz is pretty small so it's not hard to give the homeless some shelter. I hope. When I see someone begging on the streets I do assume they have somewhere to go back to.
@M.A.R. Oh, mine works in healthcare too... and is capable of seeing addicts as patients. As long as they're actually in the healthcare system, that is. As soon as they're out on the street, they're super dangerous and can't be trusted with money.
(and that goes for everyone, not just 'confirmed' addicts. Pretty much everyone she sees on the street is suspected of being one...)
@M.A.R. Can't do that here. The system is known to have too many failures, and for not being able to 'handle' everybody.
08:13
And we usually give them some money anyway, too softhearted not to. Except the ones that spray crap on your windshield. Those ones are just obnoxious
@Tinkeringbell haha yeah I dunno how older people do it. Like these modes they switch to
switch? :P
Most of them are just 'always on' in my experience XD
 
3 hours later…
10:50
@M.A.R. just another example of kids being vulnerable audience.
@Tinkeringbell lol
@M.A.R. here it's known fact the road beggars (those asking money from drivers while blocking traffic) are operated by crime organizations, they're not even beggars. And naturally, police won't do anything. But street beggars are usually people who don't really have anywhere else to be, not enough charity organizations to treat them all.
11:30
huh! 1242 vs. 1243!
 
2 hours later…
13:04
Erin Yepis on June 13, 2023
For this year’s developer survey, we added new questions to gain insight into the real sentiments behind this year’s surge in AI popularity. Is AI making a real impact in the way developers work or is it all hype? This article will recap some of the top insights, but check out Stack Overflow Labs for…
Erin Yepis on June 13, 2023
The tech that’s hot or not, and how work is changing.
Oh shut up
4
seconded - appalling timing again
What a Suspiciously Suitable data.
I always wanted a Suspiciously Suitable data like you.
Suspiciously Suitable data, do you want to be my friend?
Your skill in Lies Taming has increased by 0.1. It is now 120.0
If the data won't be what they want, survey will just be ignored. If asked, it was wiped away by a bug.
someone linked this article in a comment on meta.
putting that link on the blog would be more useful, so people would stop thinking that GPT is doing content analysis and not just text generation
13:21
@SPArcheon no because it was not written by ChatGPT.
At least if you have to generate content try to feed the dump to Flan/Alpaca. Those may have some hope to produce reasonable results.
@ShadowWizardStrikesBack ?
or if you have to hallucinate things go full nonsense and use AI Dungeon instead.
@SPArcheon Alpaca results are not cheap, 7 pounds for 400 gram.
@ShadowWizardStrikesBack ... really? That is the best joke you could come up -_-' ?
13:28
@SPArcheon yep
But people will go where other people go, that will go where money goes. They don't care about facts or truth. And so, SE following those people blindly into oblivion won't try to make things better.
13:48
@ShadowWizardStrikesBack yep, but the actual issue is not just that. My point is that if you actually want to use AI there are tons of other technologies that would at least give some actual value with about the same effort
@ShadowWizardStrikesBack why?
@SPArcheon did you really expect anything else from Sha
@SPArcheon medical?
@M.A.R. the pony is always hopeful.
@M.A.R. [:9536615] That message is not a report.
14:07
@ShadowWizardStrikesBack There is plenty of tech that has been built with the purpose of extracting information from documents based on questions given in natural language.
They actually work decent enough and they provide the indication of the source.
THAT could be a worthy use of AI on Stack.
Train the model against the datadump, have it scan the new user questions and output potential already available answers with references to the full post
Basically, a far improved "dupe detection".
To be fair, they did mention something in that regard, but we saw nothing so far.
And it is a pity, that could actually be the part where AI works.
Not significant enough.
more significant than ChatGPT
14:25
@SPArcheon They did... something, I guess
-6
Q: We are graduating the "Related questions using Machine Learning" experiment

Bella_BlueUpdate - the experiment has graduated! Thank you all for your valuable feedback. Moving forward, we plan to continue to refine this model and iterate on this feature. Please continue to post your observations and if you notice more relevance in the suggested related questions. The machine learnin...

yep, that is NOT what I had in mind :P
also... inb4 the statistical model starts to link management posts to strike posts because they very often mention each other.
14:45
> First shalt thou take out the Holy Pin. Then shalt thou count to three, no more, no less. Three shall be the number thou shalt count, and the number of the counting shall be three. Four shalt thou not count, neither count thou two, excepting that thou then proceed to three. Five is right out. Once the number three, being the third number, be reached, then lobbest thou thy Holy Hand Grenade of Antioch towards thy foe, who, being naughty in My sight, shall snuff it.
New article detected.
27
Q: When will StackExchange's senior leadership stop breaking their own code of conduct?

Sébastien RenauldA new article about the strike was posted on The Verge. In it, it is revealed that a statement by Philippe was issued, and claims the following: Later, Stack Overflow sent The Verge a press statement from its VP of community, Philippe Beaudette, criticizing moderators for levying “unnecessary su...

What did I said about "once is mistake, twice is stupidity, three is malice"?
15:16
@SPArcheon what did you say? ;)
Also, you should find something for four, five, and six.
I have some words if you need.
I doubt your suggestions can be posted.
16:17
from Twitter, linked by someone that shall remain anonymous.
> * People ask LLMs to write code
* LLMs recommend imports that don't actually exist
* Attackers work out what these imports' names are, and create & upload them with malicious payloads
* People using LLM-written code then auto-add malware themselves
note that someone on Meta reported an example of a generated answer that contained a non-existing method call.
We are getting close! Way to go!
16:43
@SPArcheon soon it would be Malware Exchange, or Malware Overflow site.
16:56
@Tinkeringbell any update on meta.stackexchange.com/questions/389879/… was just wondering. Dealing with the Reddit drama right now and going to diffrent servers etc.. in protest of third party apps
17:21
What update are you looking for?
If it works, it works. If no staff says it was fixed, good chance it will break again any moment.
Mods can't know if a bug was really fixed, or something else caused the bug to go away, that's essentially what some comment there says.
!!/fix all bugs
:(
 
1 hour later…
18:38
@ShadowWizardStrikesBack Thanks for the information. was hoping to get some rep when they mark it as compelete :P
18:56
How could there be more than 100% of people?
@RocketNikita I believe it's a multiple choice
@MetaAndrewT. still it is strange
GcL
GcL
19:16
@SPArcheon My favorite is when it returns code that has an error that is explicitly stated as something to avoid in the docs for the function being called.
 
3 hours later…

« first day (4696 days earlier)      last day (551 days later) »