In two weeks from now, on May 13 at 6:30pm, the amazing Eryk Salvaggio @cyberneticforests.bsky.social will come to Zurich to speak about «Embodied Generativity: Critical AI, Art & the Body» at ZHdK, presented by @zkk-zurich.bsky.social www.zentrumkuensteundkulturtheorie.ch//veranstaltu...

Most researchers don’t believe AGI is coming any time soon. But policy makers are steering policy toward AGI anyway. In this article for Tech Policy Press, a look at the distortions the AGI Frame introduces to policymaking. #ai #criticalai #aipolicy In @techpolicypress.bsky.social

A Nixon-era report warned of data-driven surveillance before AI even existed. Today, the government is linking databases & using “fraud detection” as a cover for targeting citizens, writes Eryk Salvaggio. What happens when free speech is redefined as a “threat”?

"generative AI, is built on a long-standing regime of extracting wealth from information, and a reliance on cheap labor to maximize the information it has collected. Treating generative AI as a revolutionary new tool,...

My latest for @techpolicypress.bsky.social, “Anatomy of an AI Coup.” With the takeover of the US government by tech elites underway, we must examine its goals and next steps — and how we will know if it has succeeded. www.techpolicy.press/anatomy-of-a...

Helloooooo BlueSky! 🌤️ We have just landed and will share more here soon 🥠 meanwhile find us and updates at -> aixdesign.co + linkin.bio/aixdesign-co/

Another fantastic read and lots of ‚food for thought’ by Eryk Salvaggio 👌🏻 @cyberneticforests.bsky.social ‚ChatGPT is not a decision-making technology, it is a decision-removing technology. It creates text to fill the space where a decision is needed’ mail.cyberneticforests.com/a-fork-in-th...

From 2021 — when it was still benign — a revealing bit of satire aimed at the “intellectual movement” that is currently shaping the policies of the United States.

“This study presents a structured examination of chatbots across three societal dimensions, highlighting their roles as objects of scientific research, commercial instruments, and agents of intimate interaction.” link.springer.com/article/10.1...

We are seeing a generation of AI systems built without their designers critically grappling with their role in the broader world. Could we do it some other way?

I wrote about David Lynch’s practice, in contrast to the model of creativity at the heart of the Gen AI image industry. I ask if the culture we cultivate inside of us could ever connect to Gen AI’s world of endlessly rearranged surfaces. mail.cyberneticforests.com/a-culture-of...

I was added to a “MAGA Trash” blocklist. I am not at all MAGA. While I don’t care about blocklists, this is a reminder to everyone to be wary of using blocklists from accounts they don’t trust.

This account exists but will likely be an automated announcement feed for future newsletters — for the most part I’m just @eryk.bsky.social!

New 404 Media: University of Michigan is selling hours and hours of audio recordings of office hours, study groups, and more to outside third-parties to train AI. Selling for tens of thousands of dollars, not clear if participants gave informed consent. I got sample www.404media.co/university-o...

Many moons ago I shared a talk about "modeling doubt," coding and engineering uncertainty into our digital systems, and weaving it into our pedagogy. That talk was published today, open access, in the Journal of Visual Culture. The whole new issue is great! journals.sagepub.com/toc/VCU/curr...

Hi folks! I'm designing an undergrad course on the history & ethics of archives & collecting across the liberal arts. We're talking archival silences, data ethics, repatriation, collecting and empire. I'm super excited about it, but struggling to find suitable reading. What would you assign? TIA!

Samsung now has a defense of its phone's AI-altering photos of the Moon: "There is no such thing as a real picture." Paging @cyberneticforests.bsky.social (not to mention Cindy Sherman) www.theverge.com/2024/2/2/240...

I actually read the original story for the first time right before reading this new one. Both are great. clarkesworldmagazine.com/kim_02_24/

Digging this fun historical perspective on the like button from Carina Albrecht in the Harvard Data Science Review. Thanks to @cheeflo.bsky.social for turning me onto this one.

“Seeing cars​ with no human inside move through San Francisco’s streets is eerie enough…but when I’m on my bicycle I often find myself riding alongside them, and from that vantage point you catch the ghostly spectacle of a steering wheel turning without a hand.” www.lrb.co.uk/the-paper/v4...

In November we announced a zine, "Models for Making Distance," which would collect instructions and manifestos for creative acts of algorithmic resistance. The deadline is April 15, and then we print. We would love to have your contributions! More details: www.cyberneticforests.com/zine

IMO crucial & productive distinction in @eryk.bsky.social 's newsletter today re: AI tools. I had lots of fun trying out Midjourney over the last year, partially because I professionally write on visual cultures & want to understand how things work, but also because I'm a total sucker for TOOLS 1/7

In “The making of critical data center studies,” @melhogan.bsky.social, Dustin Edwards, and I draw together threads across the humanities to critically examine the past, present, and possible futures of studying these weird structures we call data centers. journals.sagepub.com/doi/10.1177/...

Who is “the public” when it comes to “public engagement in AI”? journals.sagepub.com/doi/10.1177/...

need to start keeping a list of all the times some big supposed display of bleeding edge technology turns out to just be A Guy

Flicking through an older book, I come across this analogy: More data not only brings quantity but fundamentally alters quality.

In case you prefer watching a video over reading our "Reclaiming AI" paper (preprint available at osf.io/preprints/ps...), here you go 🙂 Enjoy! 🎬🍿 www.youtube.com/watch?v=akuJ...

Languages that a human couldn’t easily learn also can’t be easily learned by LLMs, according to this paper. Researchers “systematically altered English data with unnatural word orders and grammar rules.” They suggest that different architectures may have different outcomes. arxiv.org/abs/2401.06416

Made a table to make the categories of theoretical criteria, virtues and vices, accessible in: What makes a good theory, and how do we make a theory good? doi.org/10.31234/osf... I make the case for us to think explicitly about theories and place desired characteristics into virtues and vices. 1/

On Taylor Swift AI-porn and why i'm much more worried about the targeting of activists, everyday women and teenage girls, and why AI-porn enables a new digital swarm-version of the male gaze: goodinternet.substack.com/p/taylor-swi...

For the 2023 Adela arts festival, I wrote this essay about Gordon Pask: cyberneticist and conversationalist, and how his vision of “maverick machines” can help us think critically about today’s AI. wiki.ljudmila.org/Eryk_Salvagg...

“We investigate key social, political, and ethical assumptions … of AGI and human-level AI. We then draw on feminist, STS, and social science scholarship on the political and social character of intelligence … to defend a pluralistic, democratic, and participatory conception of the topic.”

"What is the role of image + aesthetic in revolution? Through a series of interviews w/... designers, Josh MacPhee charts the importance of revolutionary aesthetics from... the Black Panthers [to] ACT-UP [to] everyday organizing against nuclear power [and] for housing + international solidarity..."

A 61-year-old man was taken into custody when a private surveillance company misidentified him based on automated facial recognition and sent his name to police.

We would have seen many more articles like this if LLMs had emerged in 1995. Kate Hayles has a piece that's less psychoanalytic, but a bit like this in arguing that we need to understand the "Umwelt" of a model. muse.jhu.edu/article/898324

Structured like a language model — “We argue the intentional fictional projection of subjectivity onto LLMs can yield an alternate frame through which artificial intelligence behaviour, including its productions of bias and harm, can be analysed.” journals.sagepub.com/doi/10.1177/...

If you’ve ever asked yourself on a lazy Sunday, “What is the making of Critical Data Center Studies?” @melhogan.bsky.social, Dustin Edwards, and I have a fresh new article for you! It’s a state-of-the-field of humanistic data center research and would be a great addition to any syllabus! ☺️

all these organizations are "using data" to make their operations "more efficient" but it seems like the actual decisions they're making with that data are really naïve

Google study says that even slight alterations of digital images aimed at ML systems (like Nightshade) can register in human perception: “the brain informs us there’s a hint of cat about it.” deepmind.google/discover/blo...

Indiana University researchers explored the underground market for tools and services with names like BadGPT, XXXGPT, and Evil-GPT. OpenAI models are behind many of these products, which are used to generate malware, phishing campaigns, and scam websites. www.techpolicy.press/studying-bla...

The ever insightful @cyberneticforests.bsky.social exposes AI's performative side with cameos from Brenda Laurel, Cory Archangel, jodi.org and Gordon Pask's 1964 theater design with buttons for the audience to "weight" how a play is performed. (Start at 01:30) www.youtube.com/watch?v=04pl...

“In confronting the epistemology of AI and its regime of knowledge extractivism, a different technical mentality, a collective ‘counter-intelligence,’ has to be learned.” www.e-flux.com/journal/141/...

🎙️The Data Fix will start up again in February, with Anne Pasek. She is truly so brilliant so please subscribe now so you’ll catch the episode as soon as it’s released. shows.acast.com/the-data-fix

Recommend this piece by @eryk.bsky.social on @techpolicypress.bsky.social - LAION-5B, Stable Diffusion 1.5, and the Original Sin of Generative AI www.techpolicy.press/laion5b-stab...

With the discovery of child abuse imagery within AI’s training data, my new piece for Tech Policy Press suggests that policymakers aren’t listening to the right people about the risks of generative AI. #ai #policy

Hope you all have a happy new year! Here’s a look at the stuff I got up to in 2023, emphasizing things that were documented: resources like classes on AI art, writing on AI & tech policy from a creative research perspective, and art that’s online: open.substack.com/pub/cybernet...

I was one of many folks intvw’d for this piece about AI, human creativity, and what we lose when machines take us straight to the finish line, instead of engaging with our own limits & capacities. When we define creativity as a product, not a process, we lose what makes art personal — and “human.”

An open letter to Substack management from a collective of writers who think that the recent influx of Nazis on this platform is a problem and that their hands-off-approach to moderation doesn’t cut it anymore. Feel free to share and repost: goodinternet.substack.com/p/substacker...