Mastodon Feed: Posts

Boosted by glyph ("Glyph"):
yaelwrites ("Yael Grauer‪") wrote:

I’ve often been told to narrow this down, but my adaptability is one of my strengths. I’m most invested in finding the right cultural fit and doing meaningful, impactful work.

Boosted by glyph ("Glyph"):
yaelwrites ("Yael Grauer‪") wrote:

At the same time, I’m exploring new full-time or contract opportunities in:
•Security and privacy education
•OSINT and investigations
•Program and product management

Boosted by glyph ("Glyph"):
yaelwrites ("Yael Grauer‪") wrote:

I plan to stay in the digital security space, continuing my work as a collective member at Lockdown Systems, through IWMF’s Safety Ambassador Fellowship, and on my upcoming book about investigations for No Starch Press.

Boosted by glyph ("Glyph"):
yaelwrites ("Yael Grauer‪") wrote:

It was a pleasure presenting our research at ShmooCon, @defcon's @cryptovillage, CypherCon, Enigma, BSides, NGO-ISAC, HOPE, and many other conferences I’m probably forgetting.

Boosted by glyph ("Glyph"):
yaelwrites ("Yael Grauer‪") wrote:

I also had the opportunity to collaborate w/ amazing partners, including Tall Poppy, PEN America, and @eff. Together we dug into the efficacy of people-search removal services, the audio capabilities of video doorbells, & ways social media platforms could better address online abuse.

Boosted by glyph ("Glyph"):
yaelwrites ("Yael Grauer‪") wrote:

Beyond Security Planner, I’m particularly proud of my work on projects that called attention to the risks of memory unsafe programming languages, hyperbolic VPN marketing, and the racial, income and age disparities in people impacted by text messaging scams.

(More shoutouts to CR's testing team and survey team, as well as our fellows.)

Boosted by glyph ("Glyph"):
yaelwrites ("Yael Grauer‪") wrote:

We also ran two full content audits to make sure every recommendation remained accurate and up to date. (Shout-outs to Significant Other, @jefflandale, and CR's UX team and design team, among others, as well as the amazing tool Citizen Lab built from the ground up in the first place.)

Boosted by glyph ("Glyph"):
yaelwrites ("Yael Grauer‪") wrote:

I joined CR as a contractor in 2021 to help transition Security Planner from @citizenlab before I was brought in full-time. We rebuilt the site’s design based on UX testing, boosted traffic, and earned backlinks from CISA, Apple, and Amnesty International.

Boosted by glyph ("Glyph"):
yaelwrites ("Yael Grauer‪") wrote:

A bittersweet announcement: I was impacted by the layoffs at Consumer Reports, alongside a number of extremely talented colleagues.

I’m deeply grateful for the opportunity and so proud of the work we did fighting for consumers and helping people stay safer online.

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

His name is Ben Baker. What a soggy-looking piece of shit.

baldur@toot.cafe ("Baldur Bjarnason") wrote:

Ugh. Forgot that it’s the internet’s least funny day of the year tomorrow

Though given the state of the world, getting annoyed by the web’s yearly turn towards the unfunny is a bit like obsessing about your wet socks while you’re hanging onto a cliff’s edge by your fingernails.

Boosted by glyph ("Glyph"):
treyhunner ("Trey Hunner 🐍") wrote:

scripting language: a programming language that uses # to represent a comment

#Python #programming

NfNitLoop ("Cody Casterline 🏳️‍🌈") wrote:

Friend: [...] the point is to be able to anticipate or avoid negative scenarios

Me: Your call for help is coming from inside of the negative scenario. Please hold for an operator.

Boosted by fromjason ("fromjason.xyz ❤️ 💻 ✍️ 🥐 🇵🇷"):
blogdiva ("your auntifa liza 🇵🇷 🦛 🦦") wrote:

I SEE YOU.

i cherish you.

sending y’all big momma love and energy. you make me proud for living in your truth.

so, as i tell my babies every time they leave the nest: go out there, kick ass and take names.

be the wonderful you are.

#TransDayOfVisibility #HumanRights

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
coltofox@fox.yt ("Colto Fox") wrote:

Oh, isn't he just fab-u-lous, like yaaaasss!

#TummyTuesday #furry #fursuit #FursuitPhoto #FoxFurry #FoxFursuit #LupeSuits #belly #BellyFur #CuteFursuits #AustralianFurry

Attachments:

• unknown: A rotund Colto Fox (orange, white, brown and red fox fursuiter) in full fursuit doing a sassy pose in a room.

Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
scy@chaos.social wrote:

RE: https://furry.engineer/@soatok/116321104173678252

We shouldn't be asking why furry conventions do HIV/STI testing.

We should be asking why other conferences don't.

Excellent post about the reasoning behind it, and the cost involved, which might very well be "zero".

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

The president will have to do something else with his balls

https://www.cnn.com/2026/03/31/politics/judge-rules-that-white-house-ballroom-contstuction-stop

glyph ("Glyph") wrote:

Difficult day today. Sometimes everything just adds up and the weight of the world is a bit much, the carefully-managed bucket of despair just overflows all of a sudden.

This is definitely the sort of storm cloud that will pass, but, oof.

glyph ("Glyph") wrote:

RE: https://mastodonczech.cz/@honzajavorek/116325497238163948

Always fun to see ways that one's influence can move to places one might not expect.

Boosted by glyph ("Glyph"):
amethyst@n7.gg ("Amethyst 🌸") wrote:

Might be a long shot, but does anyone know a real human I could commission to write a short story in the Catalan/Valencian language? Bonus points if they are queer/adjacent.

baldur@toot.cafe ("Baldur Bjarnason") wrote:

I don't know how this will pan out. I don't even know the exact shape of the final book. But I think this might be a fun experiment for both me and my readers. It'll force me to do a better job of documenting my process and progress. And it'll show you some of the rough, nitty-gritty, parts of putting together a book like this.

baldur@toot.cafe ("Baldur Bjarnason") wrote:

"A massive Easter sale and a preview of my next book (out late 2026)"

https://www.baldurbjarnason.com/2026/easter-sale-and-the-toy-factory/

So for this Easter I'm making an experiment

Instead of trying to find time in between projects to work on my next book, I'm attempting to fund that work by offering a preview of it as a part of a Easter sale bundle

Those who buy the bundle or the preview also get to join a newsletter where I document the progress of the book over the coming months and, once it's out, a copy of the final version

baldur@toot.cafe ("Baldur Bjarnason") wrote:

A massive Easter sale and a preview of my next book (out late 2026): https://www.baldurbjarnason.com/2026/easter-sale-and-the-toy-factory/

Boosted by zkat@toot.cat ("Katerina Marchán"):
xgranade@wandering.shop ("Cassandra is only carbon now") wrote:

Reading analysis of the Claude Code leak (not reading the code itself, of course) is evidence towards what I had kind of suspected, that the whole thing is a giant magic trick not only in the straightforward LLMentalist way, but also in the sleight of hand way off making you think that this pile of regexes and JSON schema validation loops is *actually* the LLM doing LLM things.

Boosted by denschub@schub.social ("Dennis Schubert"):
foone@digipres.club ("Alice Averlong🏳️‍⚧️") wrote:

PSA:

1. If you are not silly, it is vital you become silly

2. If you are silly, you must stay silly

2. If you used to be silly but have stopped, you must make all efforts to return to silliness

zkat@toot.cat ("Katerina Marchán") wrote:

RE: https://mastodon.social/@arstechnica/116325367243779659

a proposal:

spin up a claude agent that will take this codebase in chunks as context and rewrite it "clean-room" and launder it into an open source project licensed under GPL

do you think they'll be cool with that? Probably, right?

EDIT: I spoke too soon. Someone already did it, and they're working on a Rust port: https://github.com/instructkr/claw-code

EmilyEnough@hachyderm.io ("Emily 🏳️‍🌈🏳️‍⚧️") wrote:

Outfit of the day for Trans Day of Visibility. I don’t hide my history online because my accounts have 20-30 years of history. But in person I just sort of blend into the background now. I never expected to “pass” in general, so it’d weird getting to a point where I need to flag with jewelry or clothes to signify my safety to other queer people now. 💜💕🏳️‍⚧️🏳️‍🌈

Boosted by glyph ("Glyph"):
mttaggart@infosec.exchange ("Taggart") wrote:

So listen.

All LLM "red teaming" is kind of a joke because of the impossibility of verifiably defending the space. There will always be a smarter mouse/jailbreak/prompt injection. But embedded application testing—that is, testing AI features once incorporated into a separate application—is uniquely pointless.

Most of the tooling to automate the drudgery of LLM red teaming assumes API access to the model/application, against which it will fire endless prompts and evaluate responses. But once the app is embedded in an application, that access is almost never available. What's left is direct application access—in other words, clicking your way to glory. Maybe you want to try to Computer Use your way to a solution, but odds are you'll just end up doing this manually. And so doing a less thorough job. And so defending even less of the possibility space.

LLMs are fundamentally insecurable, but if you only get to them once they're baked into another application, that's somehow even more the case.

Boosted by glyph ("Glyph"):
jalefkowit@vmst.io ("Jason Lefkowitz") wrote:

"There is a pervasive billionaire assumption that mainstream legacy journalism is 'liberal,' and there’s a big market for news that gets outside of the liberal bubble. That billionaire view is simply wrong. The nature of our current political moment is that the American right today is anti-empirical and anti-civic. That’s why when the Post or now CBS News goes searching for these underserved viewers hungering for journalism outside the liberal bubble, they fail. Because in a basic sense the current American right simply isn’t interested in news or journalism as most of us conceive it."

🎁 https://talkingpointsmemo.com/edblog/neutrality-authoritarianism-and-thoughts-on-the-cult-of-both-sides/sharetoken/698853da-696d-4e19-bef8-7a504700f5f7

#USPol

Boosted by EmilyEnough@hachyderm.io ("Emily 🏳️‍🌈🏳️‍⚧️"):
grumpasaurus@infosec.exchange ("Allan Chow") wrote:

I'm getting this impression that a lot of software developers would rather attempt to train an AI agent rather than train a person.