"Easter eggs, matte painting" matte paintings are usually epic film backdrops so (generated using Midjourney's CLIP-based text-to-image algorithm) aiweirdness.com/ai-generated-e…

How many languages can we support with Machine Translation? We train a translation model on 1000+ languages, using it to launch 24 new languages on Google Translate without any parallel data for these languages.arxiv.org/abs/2205.03983 Technical 🧵below: 1/18

Massively parallel web-crawl corpora are seldom what they seem! Incredibly important work putting a spotlight on serious corpus quality problems mainly affecting (as always) low-resource languages. 15 corpora contain 0% (!) in-language material...

excited beyond words to share that my provocation paper "Voice Assistants Have A Plurilingualism Problem" has been accepted to #cui2022! Looking forward to sparking discussion about why broadening the monolingual scope in IPA development matters 🌐

Really interesting case study to think about from the perspective of #ethNLP -- and also what linguists can contribute to society. >> theverge.com/2022/6/7/23153…

The draft programme for #cui2022 is out and we are excited to share it😁 Look forward to a line of awesome events, sessions and discussions happening on 26-28 July 2022! conversationaluserinterfaces.org/2022/programme/

Exaggerating a quantity? There's like a zillion, skrillion, fifty-eleven, 117, X, 36 million, threlve ways to do it. My MA thesis (on extracting hyperbolic numerals and quasi-numerals from parallel texts in 50 languages) is now freely accessible online! 📑 su.diva-portal.org/smash/record.j…

finished up an intense four-day boot camp with the 🇸🇪 and 🇮🇲 national Linguistics Olympiad teams in preparation for #IOL2022 on the Isle of Man - wishing both teams the best in today's competition. Now headed towards #cui2022 in Glasgow for three exciting conference days! 🗨🌐

Amazing discussion today at the CUIs as Entertainers SIG today ACM CUI! Thanks to @lmhclark Maylett27 and Allison Perrone for planning this with me! Want more resources on the topic? Check and edit this list! docs.google.com/document/d/1Vq… and listen to Robots on Typewriters!

#cui2022 opening keynote from @jhstuartsmith on a sociolinguistic perspective, highlighting the strong interdisciplinary nature of ACM CUI.

#cui2022 Posters and Demo session kicks off!

So excited to hear Amanda Kann talking about plurilingualism in CUIs, a topic that the community is really starting to focus on. And it was so nice to finally meet each other in person! #cui2022

Special shout out to our amazing general chairs Martin Halvey, @maryellenfoster.bsky.social & Jeff Dalton, and fellow steering committee members: @BCowanHCI @lmhclark @MPorcheron Heloisa Candello Minha Lee Cosmin Munteanu @HCI_Punk Stephan Schlögl! Not possible without them all! #CUI2022 ACM CUI

Heading home after three fantastic days at #cui2022 in Glasgow. Lovely finally meeting Zoom acquaintances in real life (plus many completely new faces), and getting to present my own first contribution to this very welcoming field! See you all in Eindhoven next summer 🗨🌐

En slags rock eller ett slags rock? Numera rockar språkvården loss och tillåter båda slagen! spraktidningen.se/2022/08/ett-sl…

Crosslingual Generalization through Multitask Finetuning 🌸 Demo: huggingface.co/bigscience/blo… 📜 arxiv.org/abs/2211.01786 💻github.com/bigscience-wor… We present BLOOMZ & mT0, a family of models w/ up to 176B params that follow human instructions in >100 languages zero-shot. 1/7

Members of Twitter's ethical AI team learned they had been laid off early this morning, according to tweets by those affected. Twitter's ML Ethics, Transparency & Accountability team has been responsible for studying algo decisionmaking, building AI fairness tools & more. (1/6)