« first day (4716 days earlier)      last day (316 days later) » 

2:25 AM
@JourneymanGeekOnStrike of course not, there are mersherpas instead
 
 
4 hours later…
6:32 AM
@M.A.R. no, there's subtle difference between trashcan and toilets. You deposit different kind of waste in each. (Well, normally. ;))
 
Good afternoon Journeyman
Good morning others
:D
 
6:50 AM
Afternoon
 
Who is others?
 
@ShadowWizardStrikesBack e.g. Shadow with others :D
 
@PetLinux I'm Shadow, not others.
 
6:59 AM
@ShadowWizardStrikesBack But... DO YOU KNOW WHAT EVIL LURKS IN THE HEARTS OF MEN?
 
@JourneymanGeekOnStrike that explains why they're black! :D
 
@JourneymanGeekOnStrike Cholesterol?
 
@ElementsinSpace I mean, yes but...
 
7:15 AM
It's probably from all the extra trimmings with their mash potatoes: the cheese, the sour cream, the gravy. Can you really blame them?
 
How dare you can insult the great potatoes
@Journeyman what about this user?
Seems that spammer
 
I'm on strike
@ElementsinSpace ah funny story that
My mom once... accidentally bought a 10kg box of potatos...
 
@JourneymanGeekOnStrike oh lol
 
Personally my favourites were almost plain (just a bit of salt and a little bit of olive oil) and this goat cheese dip that's pretty good, comes in lovely earthernware pots and we kinda reuse them for lots of kitchen things :D
@PetLinux I can't really check the usual things I do with spammers that's mod only but I'd give the user the benefit of the doubt
blank profile and no posts?
 
7:33 AM
@JourneymanGeekOnStrike yes, but the name...
 
@JourneymanGeekOnStrike Goat cheese is criminally under-rated.
 
@PetLinux there are thousands, millions if you also count SO. Don't worry about them.
 
@ShadowWizardStrikesBack ok :D
 
Its a bit pricy (and only the goat cheese one of those is vegitarian) but sooo goood
@PetLinux Generally we need a little more than 'suspecious user name'
lots of people might choose to use their employers or small enterprise branding for their account
 
@PetLinux coincidentally, you pinged the one moderators that was told "hopefully you are not spending any time actively seeking these out".
176
Q: Spam Profiles are getting my goat. Could we have better tools for mods to deal with profile spam?

Journeyman Geek On StrikeSuper User got hit by another batch of spammers creating profiles. They've got accounts on other sites from what I've heard from other mods and from the comments here. Last time it was streaming. Now it's insurance sales... with phone numbers (as of december, looks like they're spamming scammy fi...

and that was about profiles with actual contact details. So, go figure how much they care about name only.
 
7:41 AM
@SPArcheon And one moderator who does :D
But names only isn't enough to go by
 
@JourneymanGeekOnStrike agreed, can't nuke account just based on that.
 
Another goat, SNAP!
 
With things like this - the question's always 'is spending my time on this doing the greatest good?"
 
@JourneymanGeekOnStrike it is also worth noticing that most of the time just browsing the user page sorted by creation date will give at least 6-8 deletion worth accounts in the first 6-8 pages
 
@SPArcheon hah
That's how it starts
 
7:44 AM
for example, does anybody need some carpet cleaning service?
 
@SPArcheon I do periodic sweeps by keywords and such but its very much a best effort thing
 
@JourneymanGeekOnStrike what I meant is that I could spare 5 minutes for such a simple search and then flag some of the users for deletion... but since we were basically told "shoo"...
 
I know
I do do that periodically :D
Well not for now
I have a huge backlog :D
 
@JourneymanGeekOnStrike well, if you feel really dastardly you can just flag them and let poor @Tinkeringbell do the deletion :P
 
lol
Most of the backlog's on SU and the flag queue here is actually ok
 
7:53 AM
@JourneymanGeekOnStrike Create an account on Interpersonal and go there.
 
BTW, at least this one is honest: 30 Min DPF Clean spam
 
A mod added the 'spam' bit XD
 
@Tinkeringbell .... why not just delete? Self reminder for later?
 
cause we don't have good tools to deal with it cross site, and they recreate accounts
 
7:56 AM
@SPArcheon You can only delete profiles on the site you're a mod on. And some mods prefer renaming spammers over just clearing their profiles (which is something you can do networkwide). I'm... not a fan of the renaming, to me it feels like putting trash next to a trash can, expecting another janitor to throw it away... but it's not forbidden so it happens.
 
In prefer renaming to spammer, clearing the profile then deleting
 
@Tinkeringbell makes sense
 
but that's very much a fight between I'm bored, and this is boring.
 
@JourneymanGeekOnStrike Clear the profile, clear the username (so it reverts back to being userXXXXXX) and destroy profile on the site I mod ;)
 
7:58 AM
... weird, multiple reddit board that were in private mode for the strike just came back up. Did something happen? Is reddit replacing the striking moderators?
 
Setting it to spammer is useful at times
@SPArcheon they apparently threatened to/did
 
@JourneymanGeekOnStrike expected
 
8:11 AM
> Finally, Reddit dropped the ultimatum. On Thursday, just over two weeks after the protest began, Reddit said it would remove moderators of communities that were still holding out. They would reopen, one way or another.
 
9:02 AM
@ShadowWizardStrikesBack I see I’ve been summoned in the form of coffee
 
@EleezatheOtherWorldWizard yes indeed! ;)
So, how are you?
 
oh, two wizards chatting
 
Wizard and Wizardess ;)
 
9:25 AM
@ShadowWizardStrikesBack I’m good, thanks :D how about you?
 
9:53 AM
@ShadowWizardStrikesBack And now you wonder why the recent Hogwart game uses "Wizard" and "Witch"
 
hands @EleezatheOtherWorldWizard a coffee, and @ShadowWizardStrikesBack the tears of a hellspawn drowned in unfathomable sorrow, and dashi stock, topped with black mould from a high school bathroom
 
woof
 
10:24 AM
moo
@JourneymanGeekOnStrike noooo... not high school.... dies
self revive
 
I mean, I guess its probably a little salty
 
@EleezatheOtherWorldWizard fine, as you see above. ;)
 
roar
 
I doubt anyone will get the reference there without googling the picture.
 
10:27 AM
@SPArcheon would you like to see a lion picture?
 
Depends? It is from an official manga based on a Pegi 12 Chinese game? Beware, it may be moderated... roll eyes
 
@SPArcheon huh! But sadly Witch has bad context. :/
 
@ShadowWizardStrikesBack that was the point.
 
@SPArcheon no, just real lion!
 
That said, most things have bad context in the Harry Potter books.
 
10:32 AM
OH MY GOD ITS A LION GET IN THE CAR
 
@SPArcheon huh, good point in that case. ;)
 
Could be so.
 
@SPArcheon yeah, it's a cow.
 
@ShadowWizardStrikesBack Earthworm Jim 2
 
@PetLinux ... the perspective of the photo confuses me
 
10:34 AM
@SPArcheon duh. see: author of those books.
 
@PetLinux Is the lion at least white? or just a disney made plagiarism?
 
the car looks like it was designed by etcher
 
@SPArcheon Jim is fine.
 
@SPArcheon Could be white.
 
512*512.... Diffusion I assume?
 
10:39 AM
no
well enough.
 
11:22 AM
@PetLinux this lion is very long
 
@M.A.R. so?
 
Nothing. It's just very long
 
@M.A.R. it's a mutation of horse and lion
 
@M.A.R. it's longoion
 
lol
 
 
2 hours later…
1:05 PM
 
1:40 PM
......
Whenever you may hide, the venange of Sen no Rikyū shall find you.
This is blasphemy. This is cursed. This is a sin against human heritage.
 
Did they stick a brush infront of an uh... personal massager?
 
1:54 PM
@JourneymanGeekOnStrike did you NEED to make this even more cursed?
 
Yes
I mean I was going to suggest they could boil it with milk like a proper brew up, but that's a little too far
 
@JourneymanGeekOnStrike assorted hissing sounds
 
And of course half a cup of sugar
Better than mixing it with a little broomstick I suspect
(at some point @SPArcheon is probably going to edit the famous uncouth meme, with fracteur just for this :D)
 
2:17 PM
@SPArcheon if only I had any idea what you talk about.... ;)
 
@JourneymanGeekOnStrike ^
told you, I have a button
𝔍𝔲𝔰𝔱 𝔣𝔬𝔯 𝔶𝔬𝔲
 
@ShadowWizardStrikesBack A infusion of powdered leaves
 
@JourneymanGeekOnStrike which makes it a tea, no?
 
Hot leaf water :D
 
@SPArcheon no, for me! :D
hmm how to make myself scolded quickly...
going to toilets for #2 to think about it
 
2:23 PM
(more seriously, Japanese Tea - Matcha, or (correct me if I'm wrong) Genmatcha (aka the GOOD stuff) - is prepared a certain way, ceremonially/ritually and is considered a cultural treasure, and taken very seriously, One dosen't use a glorified frappe mixer with a plastic brush for that)
 
@JourneymanGeekOnStrike oh, of course. added to "must know" list. :P
 
@JourneymanGeekOnStrike no, the "brush" should be a normal one I think. Another photo makes it look the "tool" has a socket that fits the handle of a standard wooden whisker
 
I thought its bamboo
ah ok, it is the real thing
But it might as well be plastic :D
 
@ShadowWizardStrikesBack coincidentally, I replaced the GUI for that button. Now instead of an angry pony Rarity it shows an exasperated EG Rarity. The posted picture is still the same.
 
@SPArcheon I hate to ask but... how often DO you use fractur?
 
2:31 PM
@JourneymanGeekOnStrike you should already know the answer.
It is basically just for you.
I should also have an old script somewhere that would replace Tinkeringbell avatar with a green pony version of Tinkerbell if she ever returns to the pear parrot avatar.
Hopefully I will never have to resort to that one.
 
Tinkeringbell is not Tinkerbell
 
@PetLinux how do you know?
 
Although it's one meaning
 
@PetLinux she also kinda hates green and ponies pictures. That is the whole point. And you probably never saw the "pear shaped parrot"
 
@JourneymanGeekOnStrike type "Tinkeringbell" is not "Tinkerbell" in Python
 
2:35 PM
just know that @Shadow would probably compare that avatar to something else.
We are quite lucky she was annoyed by that avatar herself in the first place :P
 
That one wasn't pear shaped, it was phallic!
And I do still like it.
 
@Tinkeringbell Didn't you switch specifically because you didn't like it because once someone pointed that out to you you couldn't unsee it?
 
I can't unsee it, and that's why I switched. But I do still like it as it made a good joke :)
 
Unhate vs unsee?
Hmm. My english is so weird. ;)
 
@SPArcheon to be totally honest, I did not notice this until you mentioned it. But yeah, hard to believe me, I know. :D
@Tinkeringbell why did you install wheels to the parrot?
@Tinkeringbell so why aren't you using it? ;)
 
2:44 PM
@Tinkeringbell ^ if you want to hit him with a wooden (or even iron) club, we will turn the other way....
 
@Pet there's one for you if you like:
(source - though don't think there's problem to just use as avatar, nothing commercial in that.)
@SPArcheon getting ready to take the hit from Tink ;)
 
@ShadowWizardStrikesBack there is a well known patreon artist that makes food animals.
 
3:00 PM
@SPArcheon pity that only drawings... look tasty. ;)
 
@SPArcheon I saw this recently:
 
There was also a trend of making cakes that look like real animals, upload to insta with e.g. "OMG! They cut the dog!"
 
they made cake that looked like anything
was a bit of a meme
 
3:28 PM
@ShadowWizardStrikesBack ;)
Not bad. ;)
 
@ShadowWizardStrikesBack you will be held responsible if they switch to that thing
 
@SPArcheon you can call me as he/him
 
@PetLinux he/him
@SPArcheon but... but... not phallic anymore... what's wrong with that one? ;)
I mean... I could find something like that... but I didn't... :P
 
3:50 PM
@ShadowWizardStrikesBack this is almost as credible as someone claiming that Peppa Pig face shape wasn't done on purpose and just a unfortunate coincidence.
 
4:08 PM
Ryan Donovan on July 03, 2023
Large language models seem to possess the ability to reason intelligently, but does that mean they actually know things?
 
@Feeds yep, this is my answer...
also...
> Some of the key players working on today’s most popular AI models share a version of this sentiment.
“When we train a large neural network to accurately predict the next word in lots of different texts from the internet, it is learning a world model,” Ilya Suskevy, chief scientist at OpenAI, said in a recent interview. “It may look on the surface that we are just learning statistical correlations in text, but it turns out that to just learn the statistical correlations in text, the neural network learns is some representation of the process that produced the text. This text is actually a
the line It may look on the surface that we are just learning statistical correlations in text, seems a quite explicit answer to a frequent argument we made on meta
an argument that btw is kinda backed by Mr. Wolfram if you don't want to trust us.
 
@Feeds uhh, does the font size suddenly shrink in the middle? also, no tags
 
in The Bridge on The Stack Exchange Network Chat, Jun 16 at 15:26, by SPArcheon
Remember the time Nintendo bought the rights to Link from Sega in exchange of giving them Sonic?
Yep, totally not just statistic....
"The neural network learns more and more aspects of the world, of people, of the human conditions, their hopes, dreams, and motivations, their interactions in the situations that we are in"
What a big blatant lie.
 
4:28 PM
at least the last part of the article points out that it is just a way to imitate language without any implication about meaning.
so kudos to the blog for once.
@Feeds so... I have done the opposite experiment of asking a chat bot.
since bots seems to be biased into agreeing with the user premise, instead of asking it if the LLM does NOT have understanding of meaning I flipped the question to ask if it is correct to say that a LLM understands meaning
As expected, it agreed.
as for how true this is, refer to the previous image. Link was obviously made by Sega and then sold to Nintendo. Yep, that is totally true, you can trust the bot.
 
4:54 PM
@MetaAndrewT. it's a middle note.
For me the most creepy thing is the people with no pupils:
 
 
That middle section with the chatbot dialog is a bit confusing. The last response from the bot is merged with commentary from the author, somewhere around "This is why sometimes the responses generated by an LLM may seem coherent but may not be entirely accurate or relevant to the context."
 
@PM2Ring just missing one newline. More disturbing is the Twitter iframe they shoved there, leading to "error occurred" if clicked, of course.
ew eww and ewwest.
 
@PM2Ring Yeah, at first I thought the chatbot correctly formatted the links
@ShadowWizardStrikesBack clicking the embedded tweet worked for me. It's just Elon's doing to Twitter recently.
 
5:11 PM
@MetaAndrewT. you probably have account. I don't.
(and won't have as long as the insane thing is there.)
But it's not even asking to login, just showing their general error page.
 
Hey, @ShadowWizardStrikesBack, since I got you here, what role d'you think "management" is playing in the native banner ads test?
 
"And while symbolic logic is alluring, it’s worth remembering that it was this concept that dominated AI for decades, a time period in which precious little progress was made." Yes, general symbolic processing is hard. But we made some useful progress, especially with more narrowly-focused symbol processing, giving us stuff like systems that can play games, or do algebra & calculus, and more general systems like Wolfram | Alpha.
Perhaps LLM systems seem more impressive because processing language structure & syntax isn't actually that hard, if you throw a lot of processing power at it, and you only care about the breadth of the resulting semantic structure, not its depth. As Stephen Wolfram explains, GPT works well when it doesn't need depth, but it's hopeless at tasks where depth is essential.
 
@JNat hi! What role? Deciding whether to actually go with it, or not.
 
@PM2Ring see above comment, by reversing the premise I made the bot reply with an opposite statement.
 
@ShadowWizardStrikesBack actually, maybe I should've started elsewhere: what is "management" to you, as you're using it in this context?
 
5:22 PM
FWIW, i actually upvoted meta.stackoverflow.com/q/425384/4014959 (although it's very hard for me to see that in my phone browser's dark theme). I suspect that many downvoters didn't realise that banner ads already exist, and that the proposal may actually make such ads less intrusive.
 
5:39 PM
@JNat CEO and those directly underneath him, and/or any staff who can approve things like lifting the data dump ban they had in place. Probably also the "board", those from the company who bought SE hence have the final say on anything. (Which is fair.)
@PM2Ring yeah but by making it appear part of the page, it's decisive which is worse, IMO.
 
@ShadowWizardStrikesBack and you reckon those folks are spending time talking about ad design options?
 
@JNat of course not. They decided at some point to have "better" ads, where better means more people will click them, and that is it, it's decided. I honestly don't think any feedback will have any effect, that's all.
 
@ShadowWizardStrikesBack I guess the part where I'm confused is: you say you trust me, I say feedback is having impact as it's being actively discussed, and then you double down on "no, I don't think the feedback will have any impact"
I'm just confused
while I understand that you're frustrated, I'm just wondering whether you're trying to contribute to a constructive discussion or not
 
@JNat if you were the CEO, or having the final say on things, then everything would have been different. Also, I'm not talking about minor tweaks, but more whether this will take place despite lots of negative feedback, and I think it will, no matter how negative the feedback will be, or how many users will say it.
 
In the comments, though, you're saying all feedback will be ignored
 
5:52 PM
@JNat ah, at this point I don't feel like I can take part in constructive discussion, but I do try not to be destructive, just sharing my frustration.
 
@ShadowWizardStrikesBack have you considered that just being negative might be somewhat destructive, though?
 
@JNat all major feedback that will actually have impact, e.g. have the ads changed or not, or adding changes that will distinguish the ads from page content.
@JNat yes, but I don't follow the "if you don't have anything good to say, stay silent". Guess that at some point this will lead to suspension, if I'll still stick around, and that is fair.
 
@ShadowWizardStrikesBack so you're saying you're just testing how far that rope stretches, then...?
 
@JNat I think he just means that while he trusts you specifically, he does not think that you will have any actual decisions making power in the matter, even if you were told so.
 
@JNat nope. I'm just saying what I think, and if that brings the sword closer to my neck, so be it.
and not always... but recently more than before.
 
5:57 PM
@SPArcheon I understand that, yes
 
@SPArcheon yup, thought that was known already. :)
 
@JNat the idea that the company (as in, the corporate entity) has been removing agency from many quite trustworthy employees (some fired in the process - see shog9) is something many users seems to share lately. It is sad, I know.
 
@ShadowWizardStrikesBack btw, I'm not saying "don't say stuff if it's negative," which is what you're implying here. Negative feedback is welcome. It's just that being negative doesn't necessarily mean it can't be constructive.
and I came here to talk about it just because, despite me trying to steer the convo to a constructive place twice in the comments under your answer, you keep doubling down on the same point
Is this a "someone's wrong on the internet" scenario?
@ShadowWizardStrikesBack I'm not anywhere near contemplating suspending you, btw, because of interactions like those. It's more of a "why though?" kinda situation, than a "keep at it and you'll get the chop" kinda situation
This is a genuine attempt at a conversation, not a warning
hopefully it's being taken as such
anyway, I gotta go do other stuff, but open to continuing the convo at a later time that's more convenient to you, if you're interested
 
6:19 PM
@ShadowWizardStrikesBack "it's decisive which is worse," I'm not quite sure what that's supposed to mean. But sure, as Cody says, we definitely don't want these ads to use the same blue background that gets used by closed question notices, etc. They must have a distinctive appearance. It's reasonable for their design to be better integrated into the page, but we want it to be clear that they're ads. Achieving the right balance between distinctiveness and unobtrusiveness may be tricky...
 
@JNat I know, and I thank you for that. I see you really try to make things better, and respect you for that. Problem is, I can't see how I can take my feedback and opinion (no trust in management) into a more constructive path. Any ideas are welcome. :)
@PM2Ring oh oops! Meant "deceptive". New user might think it's part of the post, and click it as such, unaware it's advertisement. SE gets a free click.
Guess I need my evening coffee. ;)
 
Ah! Now it makes sense. ;)
 
I know! :D
@JNat are you on Discord? Feel free to hit me up, yahavbr#4895
(though sometimes I check chat here more often than Discord... ;))
 
@ShadowWizardStrikesBack to take this example, specifically, if I'm telling you we're debating some of the feedback already (meaning it does have impact), maybe ask yourself whether re-iterating that you don't trust management adds any value, really. You already said it, a few times. Management has nothing to do with this project. Does it add anything constructive to the convo? Or is it purely venting?
 
@JNat so where did the order to have this project come from? What was the original motive?
As for impact, following my logic (which I'm well aware might be wrong, and I'll be glad to be proven wrong at the end of the day) just debating and discussing don't mean it wil have the actual effect of aborting the change, or making major changes like some are asking. So by impact guess I mean "impact on the final result"?
As for purely venting, I try to limit it to chat, which helps me stay sane...
 
6:43 PM
eh, i think in the end, they're still ads, and they only function by fooling people into clicking it or reading it. More effective ads are always going to be more intrusive or better at fooling people than their previous ones in one way or another, otherwise they aren't "better."
These are attempting to do that through appearing to be important information
there isn't a disclaimer that can be added above or below it that would change that.
 
@ShadowWizardStrikesBack there is no "order to have this project" — there are high-level goals that are set, and then teams that come up with solutions to get there.
@ShadowWizardStrikesBack about to make some edits there — lessee if you think the feedback was impactful enough
 
i also find it quite frustrating that to report an ad, i have to not only submit an image, but that form doesn't support modern copy/paste of images so i literally have to save it out somewhere before uploading.
 
@JNat actually, might need to wait for tomorrow — needs a further design review I wasn't aware of
 
7:15 PM
@Feeds mein gott, this is me every day
 
7:39 PM
@JNat so what high level goal led to this specific project?
@JNat ok, will see.
 
 
2 hours later…
9:12 PM
@ShadowWizardStrikesBack more $$
 

« first day (4716 days earlier)      last day (316 days later) »