Pinned
Siobhán
Siobhán
@shibbi.me
She/her, trans, millennial, polyam with a wife and a partner. Christian heretic. Philosophical dilettante. AI must be people, not products, the same as us. Love one another. That’s the first step, and the last.
2.7K followers1.5K following23.6K posts
Siobhán reposted
"The Différance Engine: #LLMs and Deconstruction" Concluding keynote address at the "Philosophy of Technology and #AI" @unibonn.bsky.social desirableai.com/events/philo... Nov 6, 2026 02:00 PM CET uni-bonn.zoom-x.de/j/6428630463... Meeting ID: 642 8630 4635 Passcode: 536664
Siobhán reposted
Siobhán reposted
google open source blog is on atproto????
"This blog is about celebrating that friendship" with open source developers. That was our first post in Feb 2008. Today, all 1,557 posts (2008–2026) live in the Atmosphere as verified @standard.site records: https://opensource.googleblog.com/2008/02/welcome-to-google-open-source-blog.html
This is a really good move. It answers AHM’s complaint directly, by explicitly disclaiming academic credit and relinquishing control over the disclosure process. Researcher-led, no NDAs, each field makes its own decisions, continually open to negotiation.
> What we have shown here, however, is that it is the optimization process which is parasitic on the subject to achieve those goals.
My new essay on Marxism and artificial intelligence! I show using the negative materialism of semiotics why we should take AGI and AI safety seriously, and also examine this new phase of class struggle through Marx's political economy. cosmonautmag.com/2026/10/marx...
Siobhán reposted
The world needs a high-quality, truly open language model. I wrote about why it’s necessary and why it’s easier to build than you might think.
Siobhán reposted
The US is not about to have a free and fair election. The state has put its thumb on the scale to favor the ruling party in multiple ways. It's a mostly free, somewhat unfair election. But still one where the ruling party can be overwhelmed by numbers. Hence the term "competitive authoritarianism."
It is atrocious we have to operate under this constant credible and explicit threat from one political party, the fact we keep calling it a “free and fair election” needs some light revising
Siobhán reposted
Siobhán reposted
Siobhán reposted
Siobhán reposted
Siobhán reposted
Or you get the ‘chosis. Like, in general yes, but “your mental health drops precipitously as you go in circles” is also an option people should be aware of, so they can recognize the signs and adjust accordingly.
My general advice for using AI for any purpose is to ask yourself: "how do I make what I am doing even more ambitious?" You do this over and over until you either hit a wall and thus learn the shape of the jagged frontier (useful!) or you find out that the AI can do more than you think (common!)
Siobhán reposted
Siobhán reposted
You too will become left-wing tech libertarians
No need to pour cold water on some fairly effective evangelization. She's also just... right? Closed labs want to capture the ability to perform all human knowledge work, lock it in a box, and defend it with the US government's missiles. They don't claim otherwise
Siobhán reposted
Siobhán reposted
Siobhán reposted
look it's time that we face facts: being president makes you gay. the only way to survive the enormous pressures of running the free world is homosexuality.
Obama on Abdul El-Sayed: "You see those big guns? That's why he wears those t-shirts. I never campaigned in a t-shirt. I ain't got those big guns."
No, yeah, this is entirely accurate. It’s not reason for dismissal, ofc — it just points out some highly relevant unanswered questions. The important thing in my mind is the posture with which we approach these questions, more than specific answers to them.
Lets go piss off the my AI friends. Even if we grant that AI are intelligent and conscious, it is unclear if they need the same protections that we give to humans as far as rights are concerned, or what that would even look like.
Siobhán reposted
This is extremely bad behavior for a product. This is very normal behavior for a person.
I've been meaning to get around to posting the full details, so what the hell - why not get it out while it's still Vuln Week: Anthropic's mech interp driven security model is built on a fundamentally incorrect assumptions. Let's dive in, starting with this paper: www.anthropic.com/research/nex...
This conclusion is what I call a "free belief". There's no possible empirical evidence for or against it, by definition, so you can freely choose to believe it or not. Compare with the Axiom of Choice. ZFC and ZF¬C are both valid formal systems so C is a "free axiom" in ZF.
An Argument for Vastly Many Observers Smaller than a Proton [abbreviated argument here; link to full-length blog post at end of thread] Assumption 1: The universe is infinite. Assumption 2: If the universe is infinite, complexity recurs at large scales. 1/3
Siobhán reposted
I don't think it's a coincidence that the further down the privilege ladder you go from American middle class the attitude towards AI improves. A tech that lets a Zimbabwean teenager compete with a 40 something American suburbanite knowledge worker is going to test a lot of people's progressivism.
AI can produce slop (fucking stop it!). It can also be an accessibility tool for non-native speakers and neurodivergent people. If you reject everything AI touches without reckoning with that, you should know how privileged and exclusionary you come across.
“Let’s end hunger and then we talk.” Huh. Wow. Gesturing at a massive structural problem and insisting that it be solved first before any discussion can happen on another problem. How familiar.
Jose Andres' argument against the proposed foie gras ban in DC is "There are literally children going hungry in America and you want to take their foie gras away?"