In 2021, I developed PromptArray, which lets you muck around with the internals of GPT models. I moved on because this method doesn't work with closed-source models like GPT-3, but GPT-OSS makes it possible again. Read if you miss the weirdness of the GPT-2 era! jeffreymbinder.net?p=480

“Large language models like ChatGPT produce shallow, unoriginal ‘predictive text-y ideas’ and I worry that my students and others will increasingly believe that that’s okay—that there’s nothing better than that to aspire to.”

I do not see how we’re going to have productive academic discussions of "AI" until we stop accepting the marketing that lumps all machine learning methods & technologies into one amorphous thing called "AI"—so we get minimalist & maximalist responses that echo that marketing—we don’t have to do this

This is your annual reminder that Sun Ra made a Christmas song www.youtube.com/watch?v=HEpF...

Is it possible for numbers to get too big? We’ll tell you about a shocking new development, after this.

The Theories and Methodologies cluster on #criticalAI co-edited for PMLA by @ritaraley.bsky.social and myself is now online, open access. it consists of eight interventions by leading voices, as well as our framing essay “AI and the University as a Service.” www.cambridge.org/core/journal...

My book made it to the Chicago Manual of Style! I don't know how I'm going to top this.

If you see this, quote skeet with a famous landmark you’ve seen.

Thinking of the time when I bought two CDs from a stand at the WFMU record fair—"Momentum" by Monolake and "Jonathan Sings!" by Jonathan Richman—and the seller seemed to have trouble comprehending that the same person could want both of these albums.

I’ve been noticing a tendency among LLM enthusiasts to blame the prompt, not the model, whenever things go wrong. It’s snake oil salesman stuff: if my potion doesn’t cure your wounds, you just didn’t have enough faith. It makes the effectiveness of the technology itself unfalsifiable.

Got the Elden Ring DLC—and just like that, my acid reflux is back

This bit, and particularly the last sentence I’ve highlighted, is exactly where that “ChatGPT is Bullshit” paper would have benefited from engagement with a discipline that actually looks at questions like meaning, an intention, namely, literary theory. See critinq.wordpress.com/2023/06/26/a...

One thing I didn’t anticipate in my book is the sudden schism between “machine learning” and “AI” as professional affinities. Not only stuff like XGBoost, but even BERT-style transformers are now called “classical machine learning” by people into prompt engineering, RAG, etc.

If you see this, post something orange.

Google AI summaries are still not appearing for me, even though I went into "Search Labs" and enabled the feature. Am I blacklisted as a known nitpicker?

post a *perfect* album from the 90s that isn’t nirvana, pearl jam, soundgarden, or alice in chains

This Instagrammable library @ the OpenAI HQ is such a fascinating Benjaminian inversion: how does *automated* reproduction (or, in Ted Chiang’s words, blurry xeroxing or lossy compression) transform the cult + exhibition value of datasets + libraries, the latter staged here to enhance the aura of AI

Saw my first wild cybertruck just now. Reminds me of those things where someone makes a realistic version of a child’s drawing (with a side of gamer aesthetic).

People, one week left before the ACH2024 proposal due!

I couldn’t see the aurora from where I am, but at least I got to witness windows95man

Love what google is up to these days

New Homespring command just dropped www.nytimes.com/2024/04/03/u...

A hell of a way to start a book

We don’t talk enough about how Duolingo, as an attempt to undo the confusion of tongues that mankind was cursed with for building the Tower of Babel, is an affront to God

Now I’m feeling like a real Bluesky elder, even though I didn’t join until the Age of Alf.

My book Language and the Rise of the Algorithm was recently featured in the LSE Review of Books: blogs.lse.ac.uk/lsereviewofb...

CUNY is one of the most important institutions in New York. Instead of stupid articles about elite schools, I wish the NYTimes had cared about these colleges instead. I’ve seen first hand how they help young ppl move up in life (myself as a grad student at the GC included)!

Found the Windows 3.1 version of ChatGPT

This is your annual reminder that Sun Ra made a Christmas song www.youtube.com/watch?v=HEpF...

any file can be a csv file if it has a comma in it for example:

_Debates in the Digital Humanities 2023_ is now live and open-access on the Manifold DDH platform!! dhdebates.gc.cuny.edu/projects/deb... @laurenfklein.bsky.social & I offer deepest thanks to our contributors for their work and to @uminnpress.bsky.social for its support of open-access publishing!

I just switched back to MS Word after using Pages for a while, and I'm happy to find an old friend of mine in the File menu.

“SAY HELLO TO MY ONLY FRIEND!”

Just think of the splash ChatGPT would have made if it had a cooler sounding name.

It’s September 15—make sure you remember to TAKE OFF THAT STRAW HAT! en.wikipedia.org/wiki/Straw_H...

Just realized I never made an announcement on the socials. I'm back in the classroom! This semester I'm teaching the graduate course "Topics in Digital Humanities: Digital Literary Studies" at NYU. That means that today, I'm optimizing a stream processor by day and teaching Foucault by night.

Happy to publish in *Technology & Culture*, having read so much amazing work there over the years. Hoping History & STS will engage with the question of language models. Thanks to my wonderful co-authors. Explanatory 🧵 was here: zirk.us/@LivingWithM... muse.jhu.edu/pub/1/articl...

Today my girlfriend looked over my shoulder while I was working and was like, “‘BeanFactory.’ What DO you do all day?”

Spotted lanternflies are hoppin’

In honor of Gary Larson’s birthday, please share a Far Side that you love. This one could have been the epigraph of my dissertation.

As many of you are aware West Virginia University is currently being dismantled. The company doing this is RPK Group, education consultancy Take a look and see if your school is on their list of clients.

DH people: any good readings about the copyright issue in literary/humanities research? As a mainly pre-20th c. person, I never got too deep into this.

So it’s international cat day, then? Here’s a cat.

What Prosecraft was doing is clearly transformative, non-expressive, and had no effect on any market. The Hathi Trust case fought for a long time for the right to do cultural analytics research on post-1928 text. We need outrage for real, actual, critical dangers from generative AI, not this.

Not sure what to make of the prosecraft.io thing. It was bad text analysis (especially the definition of "passive voice") and a questionable interpretation of fair use. But I find it disconcerting that people are seeing simple statistical analyses and jumping straight to "AI! Kill it!"