Cerstin Mahlow @CerstinMahlow@qoto.org

Zuwachs in 28°: Sagt Hallo zur IMPULS von Paul Zimmermann, 1953 bei VEB Typoart erschienen. Weil einige fragten: das sind Klatschabzüge. Es geht hier weder um Akkuratesse, noch um typografisches Raffinement. Sie werden nur einmal abgezogen und dienen mir als schnelle Schriftprobe

@jamesravey https://github.com/FMInference/FlexGen is a really interesting development in that space too!

I'm now thinking that we will be running language models with a sizable portion of the capabilities of ChatGPT on our own (top of the range) mobile phones and laptops within a year or two

And in case you wanted to know one of the things that went wrong with Bing: https://twitter.com/mparakhin/status/1629162394764156929

> One vector of attack we missed initially was: write super-rude or strange statements, keep going for multiple turns, confuse the model about who said what and it starts predicting what user would say next instead of replying. Voila :-(

New language models released by Meta Research: https://research.facebook.com/publications/llama-open-and-efficient-foundation-language-models/

From the paper: "For instance, LLaMA-13B outperforms GPT-3 on most bench- marks, despite being 10× smaller. We believe that this model will help democratize the access and study of LLMs, since it can be run on a single GPU."

And what I mean by strategic ambiguity here is that OpenAI doesn't claim that they HAVE any rights, just that if they did have them, they're assigning them to you.

OpenAI's terms are now pretty clear (or at least, with the obvious strategic ambiguity) on IP which I assume would apply to both ChatGPT and DALL-E: "OpenAI hereby assigns to you all its right, title and interest in and to Output."

But this is new as of a TOS update in December, and the previous version does NOT say this - but I have a vague recollection that there were explicit copyright terms for DALL-E. Does anyone remember what/where this was?

In 1998, two Stanford kids published a paper in *Computer Networks*: "The Anatomy of a Large-Scale Hypertextual Web Search Engine," in which they wrote, "Advertising funded search engines will be inherently biased towards the advertisers and away from the needs of consumers."

https://research.google/pubs/pub334/

If you'd like an essay-formatted version of this thread to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:

https://pluralistic.net/2023/02/24/passive-income/#swiss-cheese-security

1/

If you drag an emoji family with a string size of 11 into an input with maxlength=10, one of the children will disappear.

Hui, dass ist mal eine gute Darstellung
#vermoegenssteuer

Der neueste Beitrag unserer #dghd_ki Reihe rund um #ChatGPT und generell #KI in der #Hochschullehre ist da:

Prof. Gabi Reinmann stellt die Frage „Wozu sind wir hier?“ und skizziert drei Ebenen, auf denen die #Hochschuldidaktik diskutieren kann/soll(?)

https://youtu.be/oSjdx-hadYY

Habt ein Auge auf die kommenden Termine im März. Dann mit @cspannagel am 13.03. und Ulrike Hanke am 20.03.

https://www.dghd.de/praxis/veranstaltungskalender/dghd-themenreihe-ki-in-der-hochschullehre/

Sneak-Peek of the project feature in the #LMS #OpenOlat from our development labs.

A versatile feature that can be used as a learning diary, project documentation, support for term papers or accompaniment of research and development projects.

Guten Morgen!