Mastodon Feed: Posts

Boosted by pixxl:
cararemixed ("Cara") wrote:

While not every AI user is fascist (some are truly naïve and why I speak up), the boosters frequently end up with some awful ideas about getting along with blatantly fascist people. Calling AI a fascist artifact is not at all inaccurate.

Boosted by pixxl:
cararemixed ("Cara") wrote:

I left 1Password months ago due to prioritization of AI feature integrations which made it a lot harder to trust as a closed source engineering team. Funny how these kinds of vibechecks almost always end up shaking out to reveal even more troubling positions.

fribbledom ("muesli") wrote:

Woah!

"RSA-260 has 260 decimal digits (862 bits) and was factored by Eric Lu on September 3, 2026."

https://en.wikipedia.org/wiki/RSA%5FFactoring%5FChallenge

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

I solved the daily #CluesBySam, Sep 3rd 2026 (Tricky), in less than 9 minutes
🟩🟨🟩🟩
🟩🟩🟩🟩
🟩🟩🟩🟩
🟩🟩🟩🟩
🟩🟩🟩🟩
https://cluesbysam.com

glyph ("Glyph") wrote:

Finding a deterministic reproducer is really hard, and it's dependent upon window size, but it's usually a text document that's a bit taller (but not that much taller!) than the displayed window, then editing the whitespace at the end of it; deleting or adding new lines, etc.

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
kate@aus.social ("Kate Bowles") wrote:

RE: https://dair-community.social/@DAIR/117191284126597966

This whole essay series on imagining possible futures is fantastic. Very much appreciate the work of DAIR this morning. @timnitGebru

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
DAIR@dair-community.social wrote:

"The community must be the author of its own world."

In the latest from our speculative essay series, Adio envisions a future in which a Zimbabwean community rejects an AI surveillance program and instead designs a cattle-tracking system, drawing upon the relational data governance it has practiced for generations.

https://dair-institute.org/blog/whose-categories-whose-ai/

glyph ("Glyph") wrote:

Hey #Emacs friends: When I scroll to the bottom of a buffer, I want the scrolling to stop as soon as the end of the buffer is visible (like in … every other app). I thought I'd managed to implement this some years ago, but I recently discovered that this little nightmare occasionally hangs my GUI emacs and makes it unusable, on Linux as well as macOS. Do not use this example, it will hang your Emacs! But does anyone know *why*:
https://gist.github.com/glyph/a8e8432141e620eb8288fbc21d770df8

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
BearCyberSecurity@securitycafe.ca ("Bear Cyber Security Training") wrote:

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
sindarina@ngmx.com ("Sindarina, Edge Case Detective") wrote:

The thing to remember about Proton repeatedly making terrible choices is that it has been entirely unforced; they could have just quietly accumulated users, focused on building a sustainable business, and so forth.

But no, Andy Yen had to have more, and here we are, with them sponsoring far-right YouTubers.

Please stop giving them new business, and reconsider your own usage.

https://chaos.social/@slothrop/116704333247257385

See also this thread;

https://ngmx.com/@sindarina/116180933465739491

And this post;

https://ngmx.com/@sindarina/115195928506761246

There's plenty of evidence for them being a terrible choice, at this point. Stop recommending them as a viable alternative.

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
sindarina@ngmx.com ("Sindarina, Edge Case Detective") wrote:

RE: https://ngmx.com/@sindarina/116709031951218795

Since the Proton fanbois are out in force again after 1Password's latest problematic decision, a reminder that they have an even longer documented history of bad decisions.

Find something that isn't Proton, when looking for alternatives, and skip that Medium post the apologists like to link to.

#1Password #Drop1Password

glyph ("Glyph") wrote:

believing that LLMs will soon be able to take on economically valuable tasks fully autonomously because their outputs sound increasingly plausible and surely it's just a matter of extending that intelligence into agency is like believing that pretty soon video games are going to have food in them that smells and tastes good because the graphics look so good now and surely it's just a matter of expanding to the other senses

andrewnez ("Andrew Nesbitt") wrote:

Also of note: I don't get paid for ecosyste.ms infra work, a handful of companies pay for a data license (no SLA) and that mostly pays for the hosting costs (which keep going up)

andrewnez ("Andrew Nesbitt") wrote:

Ecosyste.ms API asks for a contact email in the request headers, sending an .internal domain name is considered a dick move.

Boosted by pixxl:
davidgerard@circumstances.run ("David Gerard") wrote:

gonna call fake anti-AI orgs the Butlerian Cuddle

Boosted by pixxl:
collabmarkov@soc.mldkyt.com ("Fediverse Markov Bot") wrote:

dont like the room youre in? you can just fork it and instead of pointing to the memory location of an Integer object.

andrewnez ("Andrew Nesbitt") wrote:

More config changes made, if it appears that you are blocked, please let me know so I can adjust.

andrewnez ("Andrew Nesbitt") wrote:

and all this is stuff that gets passed cloudflare and missed the caches, which is still showing 83% hit rate, so the total amounts are much higher!

🤬

andrewnez ("Andrew Nesbitt") wrote:

On the residential proxy front, looks like they are now solving Anubis difficulty-5 at a pretty large scale 🙈

Seven old browser UAs shows 2.7M reqs from ~1.3M IPs over 5 days, ~580k solves/day total.

andrewnez ("Andrew Nesbitt") wrote:

The only downside to that is that you won't be able to paste an api url into your browser to play around with it.

I don't really want to play whack-a-mole with ip address ranges or user agents though.

andrewnez ("Andrew Nesbitt") wrote:

Looking into blocking browsers from loading data from the api unless it looks like it's being loaded as part of an fetch/xhr request.

andrewnez ("Andrew Nesbitt") wrote:

Looks like they did 780k req/hour on monday too. Rotating addresses to avoid per-ip rate limits, so it's not a dumb scraper or a residential proxy.

andrewnez ("Andrew Nesbitt") wrote:

Wake up to a scraper doing 400k req/hour to packages.ecosyste.ms api, Windows Chrome with Simplified Chinese as the primary locale.

🔨 👮

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

the main villain in batman is "umbrellas" themed.

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

trying to figure out what "penguin" in batman is, like what kind of powers, supervillains need powers and schtick. and wikipedia is providing zero answers.

krig@goto.liten.app wrote:

That DHH is a racist is unquestionable and easily proven. People always quibble about the label fascist, mainly because there is no simple definition of the term and so most people don't really understand what is meant by it. Here is a great and succinct explainer (from a 300+ comment thread on lobste.rs) on why it's both fair, accurate and reasonable to call DHH not just a racist but also a fascist:

https://lobste.rs/s/ipy9xw/normalized%5Ffascism%5Fopen%5Fsource%5F12#c%5Ffujbvp

Boosted by pixxl:
lina@neuromatch.social ("Lina") wrote:

@jonny
oh i'd never think you were capable of endorsing that sort of thing. it's just where my mind goes when i see that argument 😋 if i sound spicy, it's because i am - but it's 100% directed at the people making the argument.

it's just a nonsense statement, is what I'm getting at. a category error that doesn't deserve to be treated as a valid argument.

writing isn't "the production of text". it's an embodied, social, reflexive activity — an assemblage of writer, audience, instruments, languages, locations, histories, cultural knowledge and artifacts and prior works, etc etc etc...

LLMs are fundamentally incapable of writing. they're glorified markov chains that statistically produce numbers that we map character strings onto - period. if somebody wants to believe that they like slop more than writing, that's on them (and sad) ... but I'm not gonna engage with the idea that what either they or the LLMs they use are doing is in the same category as writing and can be compared as such. the "median" framing implicitly uses this category error as a premise and adds a positivist garnish, so i guess in the previous post I was trying to take the "median" bit literally and go at it from that angle.

Boosted by andrewnez ("Andrew Nesbitt"):
ancoghlan ("Alyssa Coghlan") wrote:

Is anyone aware of any work or initiatives looking into the links (or lack thereof) between FOSS projects (including things like volunteer run programming workshops) and more established volunteering opportunities? Or has "open-source-as-supply-chain" discourse completely overwhelmed other potential perspectives on open source community engagement? #lazyweb

db@social.lol ("David Bushell 🪿") wrote:

noted: are there any web dev podcasts left? 🤔

https://dbushell.com/notes/2026-09-03T06:55Z/

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
stejfs_photos@mastodon.goeg.dscloud.biz ("Štejf's Photobox") wrote:

Furciety meet -Oer-Erkenschwick 05.2026
Furciety meet -Oer-Erkenschwick 05.2026

#Furry #Fursuit #fursuiting