Stefanie Haacke
@shaacke.bsky.social
Reposted by Stefanie Haacke
In meinem neuen Newsletter blicke ich aus dem 29. Stock auf Berlin, übersetze Joshua Cohen und versuche, meine Behinderung politisch nutzbar zu machen.
42. gemeinsame Weltuntergangsverbesserung
Ich lasse die Luft aus aufblasbaren Skulpturen und behaupte, dass es möglich ist, seiner eigenen Verzweiflung gerecht zu werden.
steady.page
September 15, 2026 at 3:07 PM
In meinem neuen Newsletter blicke ich aus dem 29. Stock auf Berlin, übersetze Joshua Cohen und versuche, meine Behinderung politisch nutzbar zu machen.
Reposted by Stefanie Haacke
Please consider not saying that some AI models or some statistical models are good, useful, beneficial, whatever else positive, in science if it's not your research area. It's ok not to know things. It's ok not to both side things. It's ok to ask colleagues who are experts.
If you're anti AI & making claims that some of the models under the AI umbrella are ok because they seem to be serving scientific goals, then you fell for a type of hype. One from the 2010s; stats models aren't blessed by their use in science and they can erode our thinking. bsky.app/profile/oliv...
Another is:
2️⃣ "numerical point prediction is not something
we can nor should expect" from our models at all!
Something that I think is completely lost on so many people sadly, especially in modern artificial intelligence hype rhetoric.
5/
2️⃣ "numerical point prediction is not something
we can nor should expect" from our models at all!
Something that I think is completely lost on so many people sadly, especially in modern artificial intelligence hype rhetoric.
5/
July 10, 2026 at 4:54 AM
Please consider not saying that some AI models or some statistical models are good, useful, beneficial, whatever else positive, in science if it's not your research area. It's ok not to know things. It's ok not to both side things. It's ok to ask colleagues who are experts.
Reposted by Stefanie Haacke
So how about this for a "pause AI" proposal: "pause credulous AI reporting". No more media about "breakthroughs" or "rogue agent swarms" or ... UNLESS and UNTIL that media demands the actual empirical evidence and/or focuses on the motivations of & harms by these companies.
September 13, 2026 at 4:44 AM
So how about this for a "pause AI" proposal: "pause credulous AI reporting". No more media about "breakthroughs" or "rogue agent swarms" or ... UNLESS and UNTIL that media demands the actual empirical evidence and/or focuses on the motivations of & harms by these companies.
Reposted by Stefanie Haacke
I have consistently heard from students it that they use non university AI because they assume they are under surveillance in university systems. Would be curious to know the numbers on how many are using what universities provide. (But also: that $$ would pay for...a lot of writing centers!)
According to a meeting I was in, UMass estimates that if the capacity of these tools were maxed out by student, faculty, staff users, the total cost to the university would be $130,000/day, or about $47 million / year.
September 12, 2026 at 2:04 AM
I have consistently heard from students it that they use non university AI because they assume they are under surveillance in university systems. Would be curious to know the numbers on how many are using what universities provide. (But also: that $$ would pay for...a lot of writing centers!)
Reposted by Stefanie Haacke
«‹Alles ist möglich› – für Hannah Arendt kristallisiert sich darin eine totalitäre Idee des Machbaren: Wenn alles möglich ist, gelten keine Grenzen mehr – nicht die des Rechtsstaats, nicht die der Menschlichkeit, nicht die einfachsten Regeln gegenseitiger Rücksichtnahme.» Volker Pantenburg
Alles ist möglich / 06 / Rosa Mercedes / Harun Farocki Institut
Prometheische Schamlosigkeit mit Arendt, Anders und Farocki in Sachsen-Anhalt
www.harun-farocki-institut.org
September 11, 2026 at 2:15 PM
«‹Alles ist möglich› – für Hannah Arendt kristallisiert sich darin eine totalitäre Idee des Machbaren: Wenn alles möglich ist, gelten keine Grenzen mehr – nicht die des Rechtsstaats, nicht die der Menschlichkeit, nicht die einfachsten Regeln gegenseitiger Rücksichtnahme.» Volker Pantenburg
Reposted by Stefanie Haacke
I wrote more about what I mean about why I mark a difference between “intelligence” and “language production” here: techpolicy.press/the-illusion...
The Illusion of AGI, or What Language Models Can Do Without Thought
It is not simple stubbornness that LLMs are not “intelligent,” much less a form of “general” intelligence, writes Eryk Salvaggio.
techpolicy.press
September 6, 2026 at 2:26 PM
I wrote more about what I mean about why I mark a difference between “intelligence” and “language production” here: techpolicy.press/the-illusion...
Reposted by Stefanie Haacke
I find that, no matter how many ways I acknowledge LLMs etc are changing, people dismiss me as “crazy” or “unserious” because I refuse to attribute those changes to “intelligence.” I will say again, automated language production and intelligence are completely different things.
September 6, 2026 at 1:59 PM
I find that, no matter how many ways I acknowledge LLMs etc are changing, people dismiss me as “crazy” or “unserious” because I refuse to attribute those changes to “intelligence.” I will say again, automated language production and intelligence are completely different things.
Reposted by Stefanie Haacke
Maybe people who desperately want machines to be better than humans have the feeling that they are just not really good at being human? But none of us are. So it’s really a problem of having to deal with humanity‘s fallibility and our own split subjectivity. No machine will fix that - which is good💔
September 5, 2026 at 8:31 AM
Maybe people who desperately want machines to be better than humans have the feeling that they are just not really good at being human? But none of us are. So it’s really a problem of having to deal with humanity‘s fallibility and our own split subjectivity. No machine will fix that - which is good💔
Reposted by Stefanie Haacke
Sehr kluges Interview zu Möglichkeiten der Disruption oder der Normalisierung bei der #noAfD in #SachsenAnhalt
mit #Miteinander e.V.-Mitarbeiter der Arbeitsstelle Rechtsextremismus in Magdeburg (www.miteinander-ev.de) @davidbegrich.bsky.social
mit #Miteinander e.V.-Mitarbeiter der Arbeitsstelle Rechtsextremismus in Magdeburg (www.miteinander-ev.de) @davidbegrich.bsky.social
Ein sehr gutes Interview zur #noAfD in #SachsenAnhalt mit @davidbegrich.bsky.social. Wer ein bisschen mehr als die typischen Schlagzeilen möchte, sollte sich das unbedingt anschauen. #LTW
Wie sich die AfD in Sachsen-Anhalt radikalisierte | David Begrich im Interview | Analyse
YouTube video by Vue Critique
m.youtube.com
September 1, 2026 at 9:30 AM
Sehr kluges Interview zu Möglichkeiten der Disruption oder der Normalisierung bei der #noAfD in #SachsenAnhalt
mit #Miteinander e.V.-Mitarbeiter der Arbeitsstelle Rechtsextremismus in Magdeburg (www.miteinander-ev.de) @davidbegrich.bsky.social
mit #Miteinander e.V.-Mitarbeiter der Arbeitsstelle Rechtsextremismus in Magdeburg (www.miteinander-ev.de) @davidbegrich.bsky.social
Reposted by Stefanie Haacke
Overshoot-Szenario. Die UN beschreiben jetzt erstmals in einem Bericht offiziell das Szenario, wie die Weltgemeinschaft wieder unter 1,5 Grad kommen kann. Der Weg dahin wird härter, weil Emissionsminderung alleine nicht mehr reicht.
Hä, was?
Hä, was?
September 2, 2026 at 4:37 AM
Overshoot-Szenario. Die UN beschreiben jetzt erstmals in einem Bericht offiziell das Szenario, wie die Weltgemeinschaft wieder unter 1,5 Grad kommen kann. Der Weg dahin wird härter, weil Emissionsminderung alleine nicht mehr reicht.
Hä, was?
Hä, was?
Reposted by Stefanie Haacke
Es gibt viele Gründe "KI" abzuschalten: Extraktivismus, Neokolonialismus, Lohndruck, Deskilling, Verdummung uvm. Für die @taz.de habe ich einen anderen Aspekt fokussiert: Die entmenschlichende Gewalt, die in jeder Musterreproduktion liegt.
taz.de/Denken-wie-d...
taz.de/Denken-wie-d...
Denken wie das verbreitete Mittelmaß
KI ist fundamental unvereinbar mit emanzipierten, egalitären und freiheitlichen Gesellschaften. Dabei handelt sich nicht um einen Designfehler, sondern sie ist so gedacht
taz.de
June 20, 2026 at 7:54 AM
Es gibt viele Gründe "KI" abzuschalten: Extraktivismus, Neokolonialismus, Lohndruck, Deskilling, Verdummung uvm. Für die @taz.de habe ich einen anderen Aspekt fokussiert: Die entmenschlichende Gewalt, die in jeder Musterreproduktion liegt.
taz.de/Denken-wie-d...
taz.de/Denken-wie-d...
Reposted by Stefanie Haacke
Die derzeit in den Medien hochgejazzte Debatte zum Ost-West-Konflikt behauptet, eine Realität zu beschreiben, die medial inszeniert und zum Teil erzeugt wird. In den sozialwiss. Umfragen ist dieser Konflikt nachrangig, hinter vielen anderen, unter anderem zwischen Arm und Reich.
August 29, 2026 at 11:02 AM
Die derzeit in den Medien hochgejazzte Debatte zum Ost-West-Konflikt behauptet, eine Realität zu beschreiben, die medial inszeniert und zum Teil erzeugt wird. In den sozialwiss. Umfragen ist dieser Konflikt nachrangig, hinter vielen anderen, unter anderem zwischen Arm und Reich.
Reposted by Stefanie Haacke
»Sollte am 6. September die Stunde der AfD schlagen, wird es auch die Stunde der Reichelts, Poschardts, Döpfners und Co. gewesen sein. Sie alle haben die AfD erst salonfähig gemacht – indem sie deren rechte Narrative bedienen und damit verharmlosen.« www.blaetter.de/ausgabe/2026...
Halbloyale Demokraten: Wie die Radikalität der AfD normalisiert wird
Am 6. September, in gut sechs Wochen, findet die wohl wichtigste deutsche Wahl dieses Jahres statt, nämlich die Landtagswahl in Sachsen-Anhalt. Seit Monaten liegt die AfD in allen Umfragen klar vorn,…
www.blaetter.de
August 28, 2026 at 2:01 PM
»Sollte am 6. September die Stunde der AfD schlagen, wird es auch die Stunde der Reichelts, Poschardts, Döpfners und Co. gewesen sein. Sie alle haben die AfD erst salonfähig gemacht – indem sie deren rechte Narrative bedienen und damit verharmlosen.« www.blaetter.de/ausgabe/2026...
Reposted by Stefanie Haacke
I have seven invitations from student groups to talk about A.I.
They know what I think about A.I. That's why they asked me.
They aren't feeling it, y'all.
They know what I think about A.I. That's why they asked me.
They aren't feeling it, y'all.
At the center of my message to faculty yesterday was: 1.) We may not be adequately prepared for how angry students are about tech. 2.) They are not all angry for the same reasons. 3.) Being angry at it doesn’t mean they aren’t using it. The sense of entrapment fuels the anger.
August 27, 2026 at 10:49 PM
I have seven invitations from student groups to talk about A.I.
They know what I think about A.I. That's why they asked me.
They aren't feeling it, y'all.
They know what I think about A.I. That's why they asked me.
They aren't feeling it, y'all.
Reposted by Stefanie Haacke
„Man muss die Menschen mitnehmen.“
Nur wohin? Wohin?
Nur wohin? Wohin?
August 25, 2026 at 5:44 AM
„Man muss die Menschen mitnehmen.“
Nur wohin? Wohin?
Nur wohin? Wohin?
Reposted by Stefanie Haacke
Wild that this shit is still going on -- university administrators just plowing money into AI, with no particular vision or goal, against the wishes of most professors & students, even as money is desperately needed for other things.
Calif. colleges bought everyone ChatGPT. No one knows what to do with it.
California professors say they are left improvising as the CSU partners with OpenAI.
www.sfgate.com
August 24, 2026 at 8:21 PM
Wild that this shit is still going on -- university administrators just plowing money into AI, with no particular vision or goal, against the wishes of most professors & students, even as money is desperately needed for other things.
Reposted by Stefanie Haacke
I was kind of responding to everything from everyone in the thread since my last post.
The core issue I think is that the people training them have a flawed concept of them. I can't think of another time where engineers were so mistaken about the ontology of the thing that they themselves built.
The core issue I think is that the people training them have a flawed concept of them. I can't think of another time where engineers were so mistaken about the ontology of the thing that they themselves built.
August 23, 2026 at 3:18 PM
I was kind of responding to everything from everyone in the thread since my last post.
The core issue I think is that the people training them have a flawed concept of them. I can't think of another time where engineers were so mistaken about the ontology of the thing that they themselves built.
The core issue I think is that the people training them have a flawed concept of them. I can't think of another time where engineers were so mistaken about the ontology of the thing that they themselves built.
Reposted by Stefanie Haacke
August 7, 2026 at 5:39 PM
Reposted by Stefanie Haacke
This sign is a great example for helping to explain why chatbots present such a potent illusion.
A probably not short thread 🧵>>
A probably not short thread 🧵>>
![Photo of a sign with a doorbell attached. The printed sign reads:
CURRENT GUGGENHEIM ACCES BY HUSKY CARD ONLY
[Logos for Amazon, FedEx, UPS]
DELIVERIES/VISITORS
PLEASE RING THIS DOORBELL
Delivery hours M-F 9:00am-(obscured)
University of Washington
Taped next to the doorbell is a hand written note that reads:
yes! it rang! Pelase wait 30 seconds before pressing again.
thanks!](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:aeug3w4mdhnc2wv73hmrh36v/bafkreiapsjhnxrhqjrux3pi4np3xrwg4twz7tcrlwkmyojstelb2ndfzdi)
August 20, 2026 at 4:09 PM
This sign is a great example for helping to explain why chatbots present such a potent illusion.
A probably not short thread 🧵>>
A probably not short thread 🧵>>
Reposted by Stefanie Haacke
Woke up, made coffee, and sat at my laptop where companies gathered and shared information about my news consumption, entertainment preferences, shopping habits, and more. When I left the house, my departure time was recorded by my smart lock and my doorbell camera captured what I was wearing. 🧵
August 21, 2026 at 12:37 PM
Woke up, made coffee, and sat at my laptop where companies gathered and shared information about my news consumption, entertainment preferences, shopping habits, and more. When I left the house, my departure time was recorded by my smart lock and my doorbell camera captured what I was wearing. 🧵
Reposted by Stefanie Haacke
AI is alarming for many reasons: it's a dangerous financial bubble, an environmental catastrophe, and a tool for eroding wages and labor power. But in addition to all that, AI is an epistemic disaster.
1/
1/
August 20, 2026 at 11:52 PM
AI is alarming for many reasons: it's a dangerous financial bubble, an environmental catastrophe, and a tool for eroding wages and labor power. But in addition to all that, AI is an epistemic disaster.
1/
1/
Reposted by Stefanie Haacke
«At the ideological centre of the network is Aporia, an online magazine which serves as the public-facing arm of the Human Diversity Foundation – … a rebranded and relaunched version of the Nazi-era Pioneer Fund, which sought to mainstream its eugenics doctrines through Aporia.»
August 19, 2026 at 5:18 AM
«At the ideological centre of the network is Aporia, an online magazine which serves as the public-facing arm of the Human Diversity Foundation – … a rebranded and relaunched version of the Nazi-era Pioneer Fund, which sought to mainstream its eugenics doctrines through Aporia.»
Reposted by Stefanie Haacke
“When we say “AI models went rogue,” we skip the entire story: the part where OpenAI manually removed the model's cybersecurity blocks. We skip that OpenAI chose to test it on a machine with a live network connection.”
“Nobody is losing control of an emerging intelligence. They’re redirecting their own agency to the thing they are building.” It’s a myth, a self-affirming form of sense making: they take the action to build it, but insist on describing it as if they are passive witnesses to an inevitable force.
The System From Nowhere
When Accountability Goes Rogue
💡I recently spoke with Dr. Zena Assaad for her podcast, Responsible Bytes. Video is above, below is an essay adapted from the conversation.
"The system from nowhere" i...
mail.cyberneticforests.com
August 17, 2026 at 11:51 AM
“When we say “AI models went rogue,” we skip the entire story: the part where OpenAI manually removed the model's cybersecurity blocks. We skip that OpenAI chose to test it on a machine with a live network connection.”
Reposted by Stefanie Haacke
1. Recently I’ve received lots of emails from people I don’t know personally, asking for my feedback on their books or papers. They explain the connection between my work and theirs and ask me to take a look.
The catch, of course, is that the emails are transparently, obviously, written by AI.
The catch, of course, is that the emails are transparently, obviously, written by AI.
August 17, 2026 at 8:25 AM
1. Recently I’ve received lots of emails from people I don’t know personally, asking for my feedback on their books or papers. They explain the connection between my work and theirs and ask me to take a look.
The catch, of course, is that the emails are transparently, obviously, written by AI.
The catch, of course, is that the emails are transparently, obviously, written by AI.
Reposted by Stefanie Haacke
Dangerous referencing like this is endemic to heavily AI-assisted peer reviewed papers. Allowing the publication of LLM-generated scholarslop (to use @berrydm.bsky.social's term) is destroying science. It's an epistemic carcinogen, as @glenberman.bsky.social wrote. doi.org/10.1007/s001...
Client Challenge
doi.org
August 14, 2026 at 8:36 AM
Dangerous referencing like this is endemic to heavily AI-assisted peer reviewed papers. Allowing the publication of LLM-generated scholarslop (to use @berrydm.bsky.social's term) is destroying science. It's an epistemic carcinogen, as @glenberman.bsky.social wrote. doi.org/10.1007/s001...