Mastodon Feed: Posts

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
litchralee_v6@ipv6.social ("Litchralee_v6") wrote:

@soatok This 2022 article by Mike Masnick in the aftermath of the Twitter acquisition is usually the first thing I'll link to, on the topic of "why do we even need moderation?": https://www.techdirt.com/2022/11/02/hey-elon-let-me-help-you-speed-run-the-content-moderation-learning-curve/

I like this one because it lays out very clear that there's no "end game" for moderation at scale: it just doesn't ever let up.

Boosted by ireneista@irenes.space ("Irenes (many)"):
Cdespinosa ("Chris Espinosa") wrote:

You know what’s likely to end human life on Earth? A 3°–4° rise in global mean temperature. Let’s panic about that.

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
Firr wrote:

Doodling in my supernote is really fun. I get why @karpour loves e-ink tablets for art

Attachments:

• video: 74444d713bb2ce72.mp4

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
daemonspudguy@app.wafrn.net ("Jacob Alexander Tice") wrote:

@soatok@furry.engineer

At N > 100,000,000, God help you.

And yet every time I dare mention this to someone as an explanation (not necessarily a justification) for why moderation on massive platforms is lacking I still get the same "well they should just do it better then" nonsense as if "doing it better" is easy at a scale as ridiculous as that.

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
bbaugh@infosec.exchange ("Brian") wrote:

@soatok I already get Signal spam. Anything they can do to limit it the better. It sucks that we have no way to authenticate “this human” but having that we would have other problems. “I can’t create an anonymous account” is “I want to allow mass spam bots!”

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
monadic_cat@hachyderm.io ("Monadic Cat") wrote:

@soatok the screech I was about to write in your replies reminded me of this comic: http://www.smbc-comics.com/comic/marine-biology

speaking as a moderator for a large discord server, which itself pales in comparison to how big signal is

also, um. paying three dollars for an app is, like... normal? like that's not a big price to pay for something as useful as signal is.

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
mttaggart@infosec.exchange ("Taggart :ifin:") wrote:

The unsealed motion for summary judgment in the NYT/OpenAI/Microsoft case is a banger.

https://arstechnica.com/tech-policy/2026/09/microsoft-exec-called-ai-scraping-the-largest-theft-of-labor-in-human-history

Full document: https://cdn.arstechnica.net/wp-content/uploads/2026/09/News-orgs-v-OpenAI-Microsoft-Memo-9-17-26.pdf

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
mttaggart@infosec.exchange ("Taggart :ifin:") wrote:

Not a lawyer, but this does seem to basically wreck the Fair Use argument, and fairly demonstrate not only infringement, but knowledge of infringement.

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
chris__martin@functional.cafe ("Chris Martin") wrote:

We've decided that public job postings are more or less unusable now, because they just get flooded with AI applications. So, get #FediHired - US remote senior engineer position on my team. Full stack haskell+typescript working on practice products for k-12 students, salary range 100-150. Happy to chat or get on a phone call to discuss.

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
tante@tldr.nettime.org wrote:

RE: https://infosec.exchange/@mttaggart/117288549065128153

"Perhaps most explosively, Microsoft Director of Applied Science Brent Hecht repeatedly warned in documents that scraping news for AI training was “an astonishing theft of unprecedented proportions,” calling it perhaps the “largest theft of labor in human history,” news orgs said. In another document, Hecht contradicted Microsoft and OpenAI’s argument that training AI on news content is fair use, suggesting that the plan to widely scrape news made “a complete mockery of the idea of ‘fair use.’”"

Let him cook!

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
Yuvalne@433.world ("Talya (she/her) 🏳️‍⚧️✡️") wrote:

@soatok the introduction of phone number privacy seems to have improved things somewhat, but spam remained an issue even afterwards.
https://community.signalusers.org/t/i-thought-signal-was-spam-free/63094/2?u=rassilon1963
this thread for example talks about the need for spam detection to be improved:
https://community.signalusers.org/t/can-group-join-credentials-be-made-unique-between-two-people-to-prevent-spam/43934/5?u=rassilon1963

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
wordshaper@weatherishappening.network ("Dan Sugalski") wrote:

@soatok every time I see someone who’s entire moderation experience is running a private discord server for their five friends opine on what services handling hundreds of thousands (or millions) of people should do I just have to laugh, mostly because the alternative involves plane tickets and whacking people about the head and shoulders with a pool noodle.

Not that the people handling the large services necessarily get it right, but at least they’re actually aware of all the wrong things.

chipotle@mstdn.social ("Watts Martin") wrote:

Bevel also is convinced I don’t get enough sleep, which neither Apple Health nor Sleep++ seem to be in agreement with. It is possible I will address this problem by deleting Bevel.

chipotle@mstdn.social ("Watts Martin") wrote:

It seems like the Fitbit doesn’t record as many steps as the Apple Watch, and maybe not as much in general? Bevel, another health app I’m trying, was happy with me the last few days but today it thinks I’m a complete slacker.

soatok@furry.engineer ("Soatok Dreamseeker") wrote:

These are not scientific numbers. The actual N -> mechanism requirement will vary based on variables I can't account for.

Signal's scale is not the same as an XMPP server operated by pedantic Germans who hate UX.

soatok@furry.engineer ("Soatok Dreamseeker") wrote:

Like, there are tiers here.

A group chat with N ~ 10 people is easy to manage.

At N ~ 100 users, you typically want multiple moderators.

At N ~ 1000, you need a moderation team.

By the time you hit N ~ 10,000 you need to have runbooks for CSAM and harassment campaigns.

At N ~ 100,000 you need to seriously consider investing in automation if you want any hope of keeping abuse manageable.

At N > 100,000,000, God help you.

soatok@furry.engineer ("Soatok Dreamseeker") wrote:

Signal finally launched a means to register without giving them a phone number and techies are responding with bad faith paranoia about the registration fee.

God damn this is irritating.

Can someone who's not me and has experience dealing with spam and content moderation at scale explain how bad the problem really is? Because the dolts spreading FUD have clearly never had to manage a popular platform in their lives.

Boosted by ireneista@irenes.space ("Irenes (many)"):
owen@mastodon.transneptune.net ("Owen") wrote:

@aizuchi @mhoye if I can put mike's hat on for a second: even if you choose not to believe in the state, the state very much believes in you. That's the flip side of my take and one of the reasons I find the whole thing so frustrating.

Boosted by ireneista@irenes.space ("Irenes (many)"):
aizuchi@hachyderm.io ("Danielle Crawford") wrote:

@owen @mhoye i think the alternative, in an anarchist mindset, is that you give it away and do not concern yourself with legal enforcement, but with social enforcement. If you give it away well, there’s no financial incentive for others to exploit your work.
That’s the dream, of course. Capitalists have no incentive to borrow from mutual aid organizations. But software is more granular. In the absence of legal recourse (which hasn’t stopped any company anyway), what techniques do you think could poison-pill their profit motives?

Boosted by ireneista@irenes.space ("Irenes (many)"):
owen@mastodon.transneptune.net ("Owen") wrote:

@mhoye I mean, that's the infuriating thing, IMO: even taking the FSF's goals in the best possible light, the methods they have selected to achieve those goals depend intimately on having the resources to effectively access the legal system, in the hopes that the state can be made to step in to further those goals as part of resolving a contract dispute.

Boosted by ireneista@irenes.space ("Irenes (many)"):
owen@mastodon.transneptune.net ("Owen") wrote:

I don't think it's possible to have a license-centric view of free/open software that doesn't center the state.

Boosted by ireneista@irenes.space ("Irenes (many)"):
johnzajac@dice.camp ("John") wrote:

We really do live during the era in which "haha we tricked you!" is literally the core business model of like at least 80% of companies

Boosted by glyph ("Glyph"):
sethmlarson ("Seth Larson") wrote:

RE: https://social.rust-lang.org/@rust/117288095334006877

The lovely folks from @rust gave me a shoulder-tap in one of our mutual security discussion channels about these ongoing attacks:

https://blog.rust-lang.org/2026/09/17/targeted-attacks/

There's no particular reason the attackers wouldn't do the same to maintainers of Python projects. Please remain vigilant and send anything strange to security@python.org so we can alert others if needed.

#python #security #supplychain #pypi

Boosted by glyph ("Glyph"):
AlSweigart ("Al Sweigart") wrote:

Did you know that champagne can only be called champagne if it's from the Champagne region of France? Otherwise it is "sparkling wine".

(Trying to get a feel for how well known this is.)

Boosted by cstanhope@social.coop ("Your weary 'net denizen"):
owl@bog.fan wrote:

Duck duck go: How to be bored on an earth filled with beautiful creatures and ceaseless wonder which includes libraries

Boosted by timnitGebru@dair-community.social ("Timnit Gebru (she/her)."):
DAIR@dair-community.social wrote:

@timnitGebru spoke with Zach Stafford and Saeed Jones of Vibe Check about how we can take back the future from unimaginative AI billionaires, plus that one time she confronted Elon Musk face to face... 👀

the conversation starts at 18:40:
https://link.podtrac.com/vibecheck

Boosted by timnitGebru@dair-community.social ("Timnit Gebru (she/her)."):
DAIR@dair-community.social wrote:

DAIR's Lead Research Engineer @dylan weighed in on the mythical "kill switch" for Scientific American. Mitigating the harms of AI, as with any industry or technology, will instead require robust external regulation and safeguards.

full article (paywalled): https://www.scientificamerican.com/article/what-would-an-ai-kill-switch-do/

soatok@furry.engineer ("Soatok Dreamseeker") wrote:

Boosting a thing I wrote on Lemmy (pawb.social) because the XMPP / OMEMO crowd has weird talking points worth shining some sunlight onto:

https://pawb.social/post/50135785/24340458

jscalzi@threads.net ("John Scalzi") wrote:

COME ON THE BOAT IT IS A MAGICAL PLACE

Also I will be DJing the Nerd Prom

RE: https://www.threads.com/@jococruise/post/DdZnZOLo-6v

cwebber@social.coop ("Christine Lemmer-Webber") wrote:

Did you know that @kevingranade is on the fediverse and that she made a really cool open source videogame you can play on your computer that is allegedly about zombies but is actually about sewing and cooking and planting a garden and fixing cars https://cataclysmdda.org/