Mastodon Feed: Posts

Boosted by brib@bribstodon.xyz ("brib :neofox_floof:​ :Nonbinary:"):
dtl@8bitorbust.info ("Dr David Mills") wrote:

More Cats!
https://www.bbc.co.uk/news/articles/c6x2zgv9rr4ro

Boosted by brib@bribstodon.xyz ("brib :neofox_floof:​ :Nonbinary:"):
penelope_rattabees@mastodon.art ("penelope rattabees 🐀 🐝") wrote:

A tired sketch cause I'm tired and wanted to sketch lmao

#art #sketch #digitalart #mastoart #practice

Boosted by brib@bribstodon.xyz ("brib :neofox_floof:​ :Nonbinary:"):
mike@chinwag.org ("Mike [SEC=OFFICIAL]") wrote:

Paging @s0 you have a call waiting on the electrical crimes courtesy phone.

brib@bribstodon.xyz ("brib :neofox_floof:​ :Nonbinary:") wrote:

morning all

glyph ("Glyph") wrote:

new bio, hopefully more to the point

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
woof@aria.dog ("Dog with Glasses Plushie :neodog_glasses: :waow: ") wrote:

what’s the worst type of millennial

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
bluefops@fox.yt ("Nev :3 :nev_owo: 🔜 NextCon") wrote:

🦊🐺🏙️
(It’s 🐺🎂 tomorrow :3 )
🐺 - @voltagewolff117
📸 - Vixel
#furry #fursuit #fursuitfriday

ireneista@irenes.space ("Irenes (many)") wrote:

wait, now it fails in a different way ><

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
fzer0@nerdculture.de ("A Light Shining In Darkness") wrote:

Good morning Fedi

ireneista@irenes.space ("Irenes (many)") wrote:

this kind of has to be a long-term side-effect of our recently-added non-local exit feature, but it's odd that it didn't manifest until now

ireneista@irenes.space ("Irenes (many)") wrote:

now we'd just very much like to know how it got that way, but that is not a thing a crash handler can tell us. hm hmmm.

ireneista@irenes.space ("Irenes (many)") wrote:

okay, we added a message "control stack pointer out of range" when it wants to print a stack trace but has no joy in that task. that at least looks like clearly-deliberate behavior to a new reader, we hope.

jonny@neuromatch.social ("jonny (nonvenomous)") wrote:

i try and stay on this side of 'doing slop as performance art' vs 'actually just doing slop,' so to avoid it looking like that, i am just going to leave more context here under cw even though i just came here for the SEO and the bit is underdeveloped from where it might actually be good once i get some SEO):

the response from "university libraries" includes a claim:

While we did start the process of creating a mascot with the art and design students, for a variety of reasons including time constraints, scale, remuneration and copyright issues, we had to pivot with this particular project. In fact, this project has gotten so big that Larry has his own family tree and backstory, all created by real people.

which is funny because using the AI illustration has just way bigger copyright problems, and the scale for a cartoon otter presumably got out of hand because of the AI. and then further citing how the cartoon otter has too much lore is especially funny to me. anyway so just riffing on how people haven't digested the "ai output is not eligible for copyright" thing when doing things like "creating a mascot" that you would typically just assume copyright for because that's how it otherwise works.

glyph ("Glyph") wrote:

this would be a lot easier if there were literally ANYWHERE I could go where I could just enjoy writing some code and stop worrying about this for even a single solitary second. even “anti-AI” spaces are under constant assault by clankers and botlickers. it might be even worse there.

glyph ("Glyph") wrote:

on the one hand I really wish I hadn’t been forced to do this multiple times on the same day, and not really in a great place right now as a result, but maybe this will meaningfully decrease my ambient anxiety now that these decisions have been made

glyph ("Glyph") wrote:

when the argument *isn’t* abbreviated it’s like 50 thousand words, so that’s also a hard sell in a casual conversation

glyph ("Glyph") wrote:

not only that but the nature of the convincing lie machines is that they lie a lot, and the lies are convincing. a lot of my argument when abbreviated condenses down to “don’t believe your lying eyes” and while I have a strong conviction that that’s *correct* it sure isn’t *persuasive*

glyph ("Glyph") wrote:

and how are they supposed to know all that stuff? there are thousands of propaganda publications that they would have to know how to debunk, and I haven’t published hardly any of my own collation, so is everybody supposed to drive themselves similarly insane? it’s an unfair expectation.

glyph ("Glyph") wrote:

this is unfair of me, I have done months of reading and research that have made it so as I’m listening to boosterism I am always mentally making a list, “false, false, motivated reasoning, debunked by later research, statistical tautology, reasoning by analogy” but I can’t interrupt every point

glyph ("Glyph") wrote:

not going to tell you to quit your job or even change your beliefs just to accommodate my feelings but it makes it really hard to be around you

glyph ("Glyph") wrote:

I am not even angry at anyone, but at a time in my life when I am dealing with serious ongoing mental and physical health problems, I can’t afford the extra spoons it takes to regulate my emotions around people uncritically drowning the software community and the world in slop. it’s too much

glyph ("Glyph") wrote:

being forced to resolve several painful contradictions in my personal life today where AI disagreements have damaged certain relationships beyond repair, at least for now

Boosted by glyph ("Glyph"):
donni ("donni saphire") wrote:

Coping about the things I can’t change, seething about the things I can

Boosted by glyph ("Glyph"):
SnoopJ@hachyderm.io wrote:

from the "don't write down the deliberate crimes" desk

«A Microsoft document recognizes that…"Our AI content strategy…will hurt the performance of our models and the entire web at the same time…that is the situation we have created…"»

ireneista@irenes.space ("Irenes (many)") wrote:

oh but one good thing, with the block-sliding thing there are no longer any string pointers in the metadata that are invalid at the time they're printed. so that's great.

Boosted by glyph ("Glyph"):
leigh@ottawa.place ("Leigh Honeywell") wrote:

The real ADHD achievement is not Inbox Zero, it is Floordrobe Zero.

Boosted by glyph ("Glyph"):
linguistory@code4lib.social ("James Truitt (he/him)") wrote:

A thing I wouldn't have expected, but am hearing from a colleague who handles reference email:

As she starts getting an increasing number of AI-generated reference/appt requests, it's becoming harder for her to gauge what level of assistance any use needs, because the slop disguises the markers of inexperienced users of archives

People are so afraid to admit they don't know things, they're making it difficult for us to assess their level of knowledge and help them appropriately

jonny@neuromatch.social ("jonny (nonvenomous)") wrote:

A Murray State library made an AI generated mascot (rather than hiring idk a design student or using one of the submissions from a student project that they actually already had), and so since the mascot is ineligible for copyright protection i made a dropship slop store as a bit and so this is me posting to get the link to the store crawled by a search engine: https://www.cafepress.com/shop/LARRYPOTTERFIELDOFFICIALMERCHANDISE

ireneista@irenes.space ("Irenes (many)") wrote:

you will recall that the crash handler is also our code, so before we fix it we also need to figure out why it's not saying anything useful...

ireneista@irenes.space ("Irenes (many)") wrote:

next up is labels, there's no way around it. ironically these are only a couple lines of code to add, we've already written it (though there will be other places it needs to touch too, once we have this bit working)

unfortunately we have a mystery crash which crashes in a way that our crash handler can't diagnose