Lightnews — Scholar-powered news

Eryk Salvaggio @eryk.bsky.social · 1h

Uni Oslo, for just a too-quick visit!

Eryk Salvaggio @eryk.bsky.social · 10h

Performing Human Movie in Oslo at Trekanten, come see us! www.hf.uio.no/imk/english/...

Filmscreening: Eryk Salvaggios Human Movie – an award-winning video essay - Department of Media and Communication

We invite you to a filmscreening and performance-lecture at Trekanten with Eryk Salvaggio, a researcher, artist, and filmmaker

www.hf.uio.no

1 1 7

Eryk Salvaggio @eryk.bsky.social · 10h

If people have ideas for what to do in Oslo, I’m all ears!

7 1

Eryk Salvaggio @eryk.bsky.social · 15h

Thank you!

1 1

Reposted by Eryk Salvaggio

Kevin Buist @kevinbuist.com · 16h

Excellent stuff from @eryk.bsky.social here: mail.cyberneticforests.com/what-machine...

1 1 4

Reposted by Eryk Salvaggio

Judith Beyer @judithbeyer.bsky.social · 1d

#ContraAI 🤖 "The decision to equate human thought with complex machine slotting ... presupposes that human expression is only and without exception the automation of grammar, that words always and without exception determine ... when they will appear."
mail.cyberneticforests.com/what-machine...

What Machines Don't Know

Imagining Language Without Imagination It's important to acknowledge that Large Language Models are complex. There's an oversimplified binary in online chatter between the dismissive characterizatio...

mail.cyberneticforests.com

2 5

Eryk Salvaggio @eryk.bsky.social · 16h

Giving away control can be compelling to some people, and I think LLMs channel that energy of “handing it over to forces beyond our control.”

1

Reposted by Eryk Salvaggio

ddɐ˙ʎʞsʞɔɐlq˙uǝʌs@ @sven.blacksky.app · 19h

Reading this fantastic essay slowly as I would drink a good bottle of wine (if I did drink), every other paragraph contains a banger:

« For the same reason that a dog can go to church but cannot be Catholic, an LLM can have a conversation but cannot participate in the conversation. »

Eryk Salvaggio @eryk.bsky.social · 2d

I've been re-acquainting myself with NNs, transformers and LLMs this week, so these are early thoughts. But I am starting to think about language without the capacity to imagine language, i.e., an LLM unable to imagine itself participating in language. mail.cyberneticforests.com/what-machine...

What Machines Don't Know

Imagining Language Without Imagination It's important to acknowledge that Large Language Models are complex. There's an oversimplified binary in online chatter between the dismissive characterizatio...

mail.cyberneticforests.com

1 2

Reposted by Eryk Salvaggio

ddɐ˙ʎʞsʞɔɐlq˙uǝʌs@ @sven.blacksky.app · 20h

Consider cliches and aphorisms, thoughtless texts and emails, and the semantic satiation of overuse: "I love you," "I miss you." How do these phrases compare, or articulate, the experience of longing for someone you love? They do not, and so they serve as markers of a sentiment that fails to fulfill what they mean to do. They fail notably in contrast to a poem written about missing somebody, which strives to find new arrangements of words to articulate an experience shared by millions but in a uniquely meaningful way. In most cases, it is the effort of finding these words, not the choice of the words themselves, that move us to embrace them.

2 4

Reposted by Eryk Salvaggio

Ketan Joshi @ketanjoshi.co · 21h

"People may be using artificial intelligence, but that doesn’t mean it’s useful"

@eryk.bsky.social in @techpolicypress.bsky.social is really great:

www.techpolicy.press/generative-a...

Generative AI’s Productivity Myth | TechPolicy.Press

People may be using artificial intelligence, but that doesn’t mean it’s useful, writes Eryk Salvaggio.

www.techpolicy.press

7 47 130

Eryk Salvaggio @eryk.bsky.social · 1d

Only skimming this but it seems like it’s integrating into one model what is done externally already, re: filtering training data, generating synthetic text, etc, but with sketchier results on accuracy than if they stood alone?

1 1

Reposted by Eryk Salvaggio

Tech Policy Press @techpolicypress.bsky.social · 1d

Studies are beginning to reveal fissures in the argument that AI boosts productivity, writes Tech Policy Press fellow @eryk.bsky.social. Nonetheless, companies and governments are pouring investments into speculative growth without ample evidence.

Generative AI’s Productivity Myth | TechPolicy.Press

People may be using artificial intelligence, but that doesn’t mean it’s useful, writes Eryk Salvaggio.

www.techpolicy.press

40 61

Eryk Salvaggio @eryk.bsky.social · 1d

New from me on productivity, AI, and the problems in how we measure it: “AI’s core function is completing goals as if they were a yes/no checkmark: having a text, rather than writing the text; having a numbered list of ideas rather than finding a good idea.” www.techpolicy.press/generative-a...

Generative AI’s Productivity Myth | TechPolicy.Press

People may be using artificial intelligence, but that doesn’t mean it’s useful, writes Eryk Salvaggio.

www.techpolicy.press

2 10 34

Reposted by Eryk Salvaggio

Justin Hendrix @justinhendrix.bsky.social · 1d

The push to maximize productivity is at the heart of the current rush to adopt AI, writes Tech Policy Press fellow Eryk Salvaggio, but that doesn’t mean it’s useful. “On a social scale, productivity gains that don’t lead to pay raises, or lead to layoffs, are not productivity gains at all.”

Generative AI’s Productivity Myth | TechPolicy.Press

People may be using artificial intelligence, but that doesn’t mean it’s useful, writes Eryk Salvaggio.

www.techpolicy.press

2 18 48

Reposted by Eryk Salvaggio

Stephen Roddy @stephenroddy.bsky.social · 1d

Great read. It’s got me thinking about all of the things machines don’t care about.

Eryk Salvaggio @eryk.bsky.social · 2d

I've been re-acquainting myself with NNs, transformers and LLMs this week, so these are early thoughts. But I am starting to think about language without the capacity to imagine language, i.e., an LLM unable to imagine itself participating in language. mail.cyberneticforests.com/what-machine...

What Machines Don't Know

Imagining Language Without Imagination It's important to acknowledge that Large Language Models are complex. There's an oversimplified binary in online chatter between the dismissive characterizatio...

mail.cyberneticforests.com

1 8

Eryk Salvaggio @eryk.bsky.social · 1d

1 2

Reposted by Eryk Salvaggio

Nils Pooker @nilspooker.bsky.social · 1d

Super Essay zur Differenzierung der LLM's in der Dichotomie zwischen den Evangelisten und Hatern. Nicht die Qualität der Ausgabe ist das Problem, sondern die Uniformität als ausschließliche Rechenoperation nach einem festen System, ohne die Kontexte individueller menschlicher Gedankengenese. 1/

Eryk Salvaggio @eryk.bsky.social · 2d

I've been re-acquainting myself with NNs, transformers and LLMs this week, so these are early thoughts. But I am starting to think about language without the capacity to imagine language, i.e., an LLM unable to imagine itself participating in language. mail.cyberneticforests.com/what-machine...

What Machines Don't Know

Imagining Language Without Imagination It's important to acknowledge that Large Language Models are complex. There's an oversimplified binary in online chatter between the dismissive characterizatio...

mail.cyberneticforests.com

1 1 9

Eryk Salvaggio @eryk.bsky.social · 1d

Thinking about AI & our "models of models," "vector space" creates a metaphor of motion that feels multidimensional. It's a helpful metaphor, and translates calculus well, so it's useful. But tracking change is not limited to motion, so "flat" can be useful too.

2 7

Reposted by Eryk Salvaggio

Roland Meyer @bildoperationen.bsky.social · 2d

In this great piece, Eryk reminds us that in order to be sceptical about so-called AI, we don’t have to dismiss LLMs as mere fancy auto-complete. As a colleague from linguistics once remarked, we overestimate AI as a simulation of thinking, but we may underestimate it as a model of language
1/

Eryk Salvaggio @eryk.bsky.social · 2d

I've been re-acquainting myself with NNs, transformers and LLMs this week, so these are early thoughts. But I am starting to think about language without the capacity to imagine language, i.e., an LLM unable to imagine itself participating in language. mail.cyberneticforests.com/what-machine...

What Machines Don't Know

Imagining Language Without Imagination It's important to acknowledge that Large Language Models are complex. There's an oversimplified binary in online chatter between the dismissive characterizatio...

mail.cyberneticforests.com

6 24 83

Eryk Salvaggio @eryk.bsky.social · 2d

Tech is full of philosophers who couldn’t get jobs in philosophy

2

Eryk Salvaggio @eryk.bsky.social · 2d

Where are the 1990s

Land of the Loops Multi-Family Garage Sale music video

YouTube video by Alan Sutherland

youtu.be

3

Reposted by Eryk Salvaggio

Simon Groth @simongroth.com · 2d

“Current architectures of LLMs cannot imagine, but they can sequence … For the same reason that a dog can go to church but a dog cannot be Catholic, an LLM can have a conversation but cannot participate in the conversation.”

Eryk Salvaggio @eryk.bsky.social · 2d

I've been re-acquainting myself with NNs, transformers and LLMs this week, so these are early thoughts. But I am starting to think about language without the capacity to imagine language, i.e., an LLM unable to imagine itself participating in language. mail.cyberneticforests.com/what-machine...

What Machines Don't Know

Imagining Language Without Imagination It's important to acknowledge that Large Language Models are complex. There's an oversimplified binary in online chatter between the dismissive characterizatio...

mail.cyberneticforests.com

1 7 30

Reposted by Eryk Salvaggio

mcc @dryad.technology · 2d

What I'm listening to today: "Olson"

Boards of Canada arranged to play on a DEC PDP-1 from 1959. The PDP-1 doesn't have sound, but it does have front-panel light bulbs for debugging, so they rewired the light bulb lines into speakers to create 4 square wave channels

www.youtube.com/watch?v=wubk...

Boards of Canada "Olson" on a 1959 PDP-1 Computer

YouTube video by Joe Lynch

www.youtube.com

5 37 95

Eryk Salvaggio @eryk.bsky.social · 2d

Recent posts re: Meta’s goals to build solipsistic networks that simulate social interaction through sycophantic LLMs to replace their dead mall vibe reminded me of this text, written in 2023, which tied the logics of social media analytics directly to logics of generative AI.

Eryk Salvaggio @eryk.bsky.social · 2d

“Now AI promises to further constrain those relationships, to move us from a time when one could speak and hear from many to a time when one can speak only to ourselves: one-to-none communication, a throwback to the days of yelling at the TV, but now the TV can adjust.” (From 2023).

The Hypothetical Image — Cybernetic Forests.

The history of AI images as data analytics, and AI art as the aestheticization of Big Data’s politics.

www.cyberneticforests.com

2 2 10

Eryk Salvaggio @eryk.bsky.social · 2d

“Now AI promises to further constrain those relationships, to move us from a time when one could speak and hear from many to a time when one can speak only to ourselves: one-to-none communication, a throwback to the days of yelling at the TV, but now the TV can adjust.” (From 2023).

The Hypothetical Image — Cybernetic Forests.

The history of AI images as data analytics, and AI art as the aestheticization of Big Data’s politics.

www.cyberneticforests.com

1 13