Mastodon Feed: Posts

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

You can ditch a place but you can't ditch your own brain

Boosted by brib@bribstodon.xyz ("brib :neofox_floof:​ :Nonbinary:"):
perfect10_bot@infosec.exchange wrote:

Vendor:
GitLab

Products: GitLab
CVE-2026-85706 - https://www.cve.org/CVERecord?id=CVE-2026-85706

... a perfect 10. you know who else is a 10? :neobot_blush_hide:

neatnik@social.lol ("Neatnik :prami:") wrote:

Thanks to everyone who shared a story on https://eph.emer.al/september-11 today. 💜 The stories will be around for a few more hours, and then the site will go back to its regularly scheduled programming of temporary thoughts.

jsonstein@masto.deoan.org ("Jeff Sonstein") wrote:

this chocolate milkshake blows out my entire month’s sugar budget

Boosted by neatnik@social.lol ("Neatnik :prami:"):
davidcelis@xoxo.zone ("david celis") wrote:

so is Ruby Central gonna report OpenAI to the FBI or...

https://rubyhack.ai/

Boosted by neatnik@social.lol ("Neatnik :prami:"):
NanoRaptor@bitbang.social ("Nanoraptor") wrote:

A sticker to place anywhere.

Like hand dryers, toilet bowls, soap dispensers, alarm keypads, coffee machines, thermostats, pedestrian crossing buttons, BBQs, park benches, parking meters...

As well as staplers, pens, potatoes, bricks, brooms, lunch, potplants and speed humps.

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

I solved the daily #CluesBySam, Sep 11th 2026 (Tricky), in less than 21 minutes
🟩🟩🟨🟩
🟩🟩🟩🟩
🟩🟡🟩🟩
🟩🟨🟡🟩
🟩🟨🟡🟩
https://cluesbysam.com

Boosted by neatnik@social.lol ("Neatnik :prami:"):
fight@fightforthefuture.org ("Fight for the Future") wrote:

We maintain that there's no way to implement age verification without compromising sensitive data. Those who choose to opt out of this get a censored experience.

A reminder: the leaders pushing for age verification broadly aren't doing so for child safety. Their aim is information control.

Boosted by brib@bribstodon.xyz ("brib :neofox_floof:​ :Nonbinary:"):
maxleibman@beige.party ("Max Leibman") wrote:

Or McDo not. There is no try.

Boosted by glyph ("Glyph"):
itamarst@hachyderm.io ("Itamar Turner-Trauring") wrote:

Human beings are very good at imitation; to some extent we can't help but imitate.

Unfortunately this means if someone exposes their brain to chatbot outputs all day, they might start writing like a chatbot without realizing it. And then get offended when people claim their writing is slop...

Boosted by glyph ("Glyph"):
devxvda@mastodon.ie ("Dr.Nick") wrote:

We are almost there folks hang in there

2020 ▓▓▓▓▓▓▓▓▓▓▓▓▓░ 669%

fromjason ("fromjason.xyz ❤️ 💻 ✍️ 🥐 🇵🇷") wrote:

RE: https://mastodon.social/@fromjason/115843118955740563

Not bad, brother.

Boosted by andrewnez ("Andrew Nesbitt"):
tenderlove ("Aaron Patterson ✅") wrote:

What a time to be alive https://tenderlovemaking.com/2026/09/11/what-a-time-to-be-alive/

cstanhope@social.coop ("Your weary 'net denizen") wrote:

The partial lyrics of two songs my brain keeps alternating between... I assume they'll eventually tire of chasing each other around in the folds of my brain...

cstanhope@social.coop ("Your weary 'net denizen") wrote:

"Fast, slow, fast, fast, slow"

cstanhope@social.coop ("Your weary 'net denizen") wrote:

"More human than human I can do it all: french kiss, smoke a spliff, give a shit about baseball"

Boosted by timnitGebru@dair-community.social ("Timnit Gebru (she/her)."):
jo@infosec.exchange ("jo the disgraced") wrote:

@DAIR I firmy believe it's a PR move; the threat posed by LLMs isn't Terminators, it's aiding mass surveillance and drowning the world in slop.

Boosted by timnitGebru@dair-community.social ("Timnit Gebru (she/her)."):
arteesetica ("Arte es Ética") wrote:

Daños reales que la industria tecnológica podría perpetuar: «que la IA impulse armas autónomas, máquinas de matar, que de hecho se están utilizando en la guerra. La catástrofe climática podría verse agravada por lo que estas empresas están desarrollando. Los jefes están utilizando la IA como excusa para deshacerse de sus "molestos" trabajadores» 🎯 @timnitGebru

https://dair-community.social/@DAIR/117253572479354319

#AI #hype #tech #technology #OpenAI #Anthropic #AGI #machine #Skynet #Coxon #JacobCoxon #security

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

The reason google never did this with e.g. google docs seems like because "just directly plagiarizing by rephrasing a document" would be outrageous scandal, but when you have a billion dollar information laundering machine you get to call it AGI

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

RE: https://mathstodon.xyz/@andreasthom/117240535270608201

We can assume unreleased models under active development are being trained and retrained constantly, that's what developing them means, so they are exposed to new data that enters the training set. The high capacities of these models mean that they are absolutely capable of reproducing unique training data nearly exactly. In a research setting like this where the model provider is itself doing the research we can also assume they can disable any filters or safeguards that might normally discourage/prevent output that looks like leaking training data in pursuit if a goal.

With all that, the distinction between "user data being entered into the generating context" and "user data being entered into the training data" is not really meaningful.

So it seems like if your research is identifiable/tagged to a named result as some of these proofs are, or really in any condition where you'd expect your work being in training data would meaningfully impact the output for a question, the question of whether an AI company swoops in to steal your work seems like more or less a "will they" as a cost/benefit per the prestige of the problem, rather than a "can they"

Boosted by jonny@neuromatch.social ("jonny (nonvenomous)"):
andreasthom@mathstodon.xyz ("Andreas Thom") wrote:

@tristanbuckmaster

3/3 Sellke’s categorical answer was, at minimum, unjustifiably broad and materially misleading; looking back it was plainly dishonest. Bubeck’s acknowledged attacks and his attempt to exclude Alpöge from a proposed OpenAI-proof paper deepen the concern that there is a loss of moral compass. Taken together, this raises serious questions about their judgment and personal integrity.

Boosted by jonny@neuromatch.social ("jonny (nonvenomous)"):
andreasthom@mathstodon.xyz ("Andreas Thom") wrote:

@tristanbuckmaster

2/3 OpenAI is drawing a distinction that its answer to me erased, despite the fact that my question explicitly made that distinction.
We are not required to reverse-engineer OpenAI’s internal training pipeline to establish what happened. Only OpenAI has the relevant data for that. If Sellke and Bubeck give a categorical denial, it must disclose the basis: product and privacy settings, relevant datasets and checkpoints, and what “de-identified data derived from usage” means.
I disabled model training on 29 June. That control is still only a promise whose implementation users cannot audit, and it is prospective: it does not answer what happened to earlier conversations or to derivatives already selected. OpenAI’s answer did not mention the setting or any account-specific check.
There is also a mathematical parallel, though a less explicit one. The approach of Kun and myself was not the main line of attack on non-soficity, in fact there were other more promising approaches along the line of quantum games etc. OpenAI’s detailed command of our techniques therefore made me wonder how the model found this route. If nonpublic research supplied by users improved a model and the provider then used that model to race those users to publication - without informed consent, disclosure, or credit - that would be ethically indefensible. De-identification may remove a name; it does not remove the intellectual content of a mathematical idea. Sellke and Bubeck seem to be blind to this simple moral aspect.

Boosted by jonny@neuromatch.social ("jonny (nonvenomous)"):
tristanbuckmaster wrote:

Today, Levent Alpöge and I have made public three results: finite-time blowup with smooth forcing for incompressible porous media, for Boussinesq, and for 3d incompressible Euler.

https://cims.nyu.edu/~tristanb/statement.pdf

https://cims.nyu.edu/~tristanb/euler.pdf

https://cims.nyu.edu/~tristanb/ipm.pdf

https://cims.nyu.edu/~tristanb/boussinesq.pdf

https://github.com/tristanbuckmaster/fluid%5Flean

Boosted by jonny@neuromatch.social ("jonny (nonvenomous)"):
andreasthom@mathstodon.xyz ("Andreas Thom") wrote:

@tristanbuckmaster

1/3 I want to record a number of points that I realized after reading up on the controversy. In fact it recalls an exchange I had with OpenAI after its non-sofic-group announcement. It exposes a serious, unanswered question about whether researchers can trust OpenAI with unpublished mathematics.

I wrote in an email to Mark Sellke and Sebastien Bubeck shortly after the surprising finding of a non-sofic group using the methods of Kun and myself:
“Another point is that I and a colleague in Dresden were discussing the expander matching problem and various extensions of the work with Gabor Kun actively over the last months with ChatGPT, so that we are of course curious if that was part of the training data or accessible to the reasoning process. There is a certain (frankly unacceptable) lack of transparency here; and I fear it will damage the communal process of math more than the new AI-generated results will benefit the subject.”

Mark Sellke’s complete answer was: “Regarding your conversations with ChatGPT: that did not happen.”

I had explicitly asked about two different things: (1) whether our conversations entered training data, and (2) whether they were accessible to the solving process. The categorical answer now looks as though it addressed only direct access under (2). No such qualification, explanation, or evidence was given. I take this as dishonesty to say the least.
OpenAI now says in the Buckmaster-Alpöge case that no specific user data was accessed, but adds that it “cannot rule out that de-identified data derived from their usage of our products helped improve our models.”

Boosted by EmilyEnough@hachyderm.io ("Emily 🏳️‍🌈🏳️‍⚧️"):
gwynnion ("Nowhere Girl") wrote:

In a very real way, the hazing and bullying was meant to push you back into the closet, to make you deny it, make you act differently to stop it, and eventually stop thinking about it.

Even some girls who knew opted to hide rather than eat shit on a daily basis for the first 18 years of their lives.

There's no shame in that.

Boosted by ireneista@irenes.space ("Irenes (many)"):
thomasfuchs@hachyderm.io ("Thomas Fuchs 🫯") wrote:

I'm not sure how to explain this, but you can't take anything a fascist says at face value.

Everything they fart into the world from their face holes and everything they ooze on to a (web) page is poison designed to gaslight you.

"But the project that he runs gives money to open source" is like saying "but Hitler was nice to dogs".

jsonstein@masto.deoan.org ("Jeff Sonstein") wrote:

when the hell did Duboce Park become a dog park??

Boosted by jwz:
dnalounge@sfba.social ("DNA Lounge") wrote:

♬️ Just announced: Thu Oct 15, 7pm: ROCKY HORROR PICTURE SHOW
https://www.dnalounge.com/calendar/2026/10-15.html?utm%5Fsource=sp%5Fma
#dnalounge #rockyhorrorpictureshow #cyberdelia #thebawdycaste #glam #sanfrancisco

Attachments:

• gifv: 10aecee54185c01a.mp4

Boosted by brib@bribstodon.xyz ("brib :neofox_floof:​ :Nonbinary:"):
lcamtuf@infosec.exchange ("lcamtuf :verified: :verified: :verified:") wrote:

One buyer confirmed!

Boosted by brib@bribstodon.xyz ("brib :neofox_floof:​ :Nonbinary:"):
eclectech@things.uk wrote:

A bit of cuteness before bed. Happy feathery fluffy dreams all, when night reaches you.

#sillyScribbles #featherFolk