Cerstin Mahlow @CerstinMahlow@qoto.org

So OpenAI just released a detector of AI-generated text, I assume because of concerns in education / homework.

https://openai.com/blog/new-ai-classifier-for-indicating-ai-written-text/

Maybe this is good?

No, it's very bad.

They claim 26% true positives, 9% false positives. Assume 10% of submitted homework is chatgpt generated, you get the classic counterintuitive outcome of poor predictive power: if a homework is flagged, there's a 3:1 chance it's *human* generated.

This is going to cause a lot of harm. It should be immediately recalled.

#whataweekhuh

"Instead of “playing” with #ChatGPT (cough, nota toy, cough) in your class you could play the Data, Privacy, and Identity game developed by Jeannie Crowley, Ed Saber, and Kenny Graves"

This is just one idea I've curated and published in "Prior to (or instead of) using ChatGPT with your students" a new post on my personal blog

https://autumm.edtech.fm/2023/01/18/prior-to-or-instead-of-using-chatgpt-with-your-students/

#EdTech #FacDev #Privacy

OK, the Friedrich-Alexander-Universität Erlangen-Nürnberg wanted to simultaneously fill "6 Open Topic W3-Professuren im Themenfeld Digital Humanities and Digital Social Studies", application deadline was 2021-05-23. Sounded bold and ambitious, so especially disappointing to learn that the appointment process has been stopped for strategic-structural reasons.
Time it took to come to this decision: one-and-a-half years, so almost not slow for a German university :)
#ichBinHanna

And in case this post wasn't clear: I'm all-in on large language models: they confidently pass my personal test for if a piece of technology is worth learning:

"Does this let me build things that I could not have built without it?"

What I find interesting is that - on the surface - they look like they solve a lot more problems than they actually do, partly thanks to the confidence with which they present themselves

Figuring out what they're genuinely good for is a very interesting challenge

Eine Wolke, die "Regen!" ruft und damit zu mehr Bewegung motiviert. Ein Feststecker, den nur ausstöpseln kann, wer auf der Leitung steht. Eine Portion Beratkartoffeln für mehr Tschakka im Leben. Feine Dösaromen fürs Abendessen. Tropfen gegen Kopfschüttelfrust: 30 Stück, 3x täglich. Ein Zeugnis voller Fußnoten (links besser als rechts).

#SechsUnmöglicheDinge

Eine künstliche Intelligenz, die schreibt – wird der Mensch als Au­to­r:in bald überflüssig sein? http://www.taz.de/!5909029/

What is this, a #fantasy #caturday crossover episode?

Das mag niemand.

#märchen #hänselundgretel #hexen #dachschaden #cartoon #comic #funny #bäcker #bäckerei #handwerk #handwerker

In roughly the past half-decade, Microsoft went from nowhere to overwhelming dominance of text editors with VSCode, ownership of majority of code hosting (and open source dev) with GitHub, ownership of the dependency stack used by most devs with npm, control over the most popular single language with TypeScript, and is trying to position copilot and ChatGPT as inevitable parts of the future dev process. Nothing negative for the ecosystem will come of this, as the last half century teaches us.

One of the things I'm finding so interesting about large language models like GPT-3 and ChatGPT is that they're pretty much the world's most impressive party trick

All they do is predict the next word based on previous context. It turns out when you scale the model above a certain size it can give the false impression of "intelligence", but that's a total fraud

It's all smoke and mirrors! The intriguing challenge is finding useful tasks you can apply them to in spite of the many, many footguns

Nüülis, an't Köhlschapp ...

Esszimmer (Beelitz, Heilstätten, Brandenburg)

"Ich habe zu schnell gehandelt. Ja. Das war's." Interview von Matthias Sander mit He Jiankui, dem chinesischen Biophysiker, der 2018 die Welt mit der Erschaffung von drei genmanipulierten Babys schockierte.
#HeJiankui #bioethics

https://www.nzz.ch/technologie/he-jiankui-erschuf-drei-genmanipulierte-babys-jetzt-bricht-er-sein-schweigen-ld.1722020

♫ it's the i of the tiger, with a g to the right
Risin' up to the challenge of a vowel
And the last two letters are an e and an r
And the t is in front of the iiiiii
of the tiger ♫

Lastwagen beendet Überholmanöver nach 14 Jahren
https://www.der-postillon.com/2013/01/lastwagen-beendet-uberholmanover-nach.html

2019: big petition in the #machineLearning community to allow remote presentations at scientific conferences

2020-21: ML conferences are virtual-only

2022: several major in-person conferences are super-spreader events

2023: "at least one author of each accepted paper must attend to present their work in person"

I know the virtual formats were less than ideal, but I can't blame junior researchers who are torn between flying halfway across the globe for a conference and harming their career.

My colleague, John F. Hughes, got asked for a paper that he hadn't written. The correspondent replied: "I used the chat.openai to gather new sources that I might have missed during my own scan of literature".

(I know librarians are being driven bonkers by GPT-3-manufactured book titles. But this is the first time I've heard of an academic paper request…)

@jesseabe For me personally (I don’t think there are many who’d agree), some foundational texts for #DigitalHumanities are:

- Granger, Gilles-Gaston (1967). Pensée formelle et sciences de l'homme. Paris: Aubier-Montaigne. [There is an English translation, which I’d avoid if you can read the original.]

- Гладкий, А. В., & Мельчук, И. А. (1969). Элементы математической лингвистики. Москва: Наука. [There are French, German, and English translations.]

- Leff, Gordon (1972). Models inherent in history. In T. Shanin (Ed.), The Rules of the Game (pp. 148–160). Abingdon: Tavistock.

- Stachowiak, Herbert (1973). Allgemeine Modelltheorie. Wien, New York: Springer.

- Gardin, Jean-Claude (1991). Le calcul et la raison. Paris: Éd. de l'École des hautes études en sciences sociales.

- Moles, Abraham A. (1995). Les sciences de l'imprécis. Paris: Seuil.

- Gardin, J. (2012). Modèles et récits. In J. Berthelot (Ed.), Épistémologie des sciences sociales (pp. 407–454). Paris: Presses Universitaires de France. https://doi.org/10.3917/puf.berth.2012.01.0407

- Meunier, Jean-Guy (2014). Humanités numériques ou computationnelles. Sens public. http://sens-public.org/articles/1121/