J. Martin

March 12, 2023

Schroedinger’s Schnelltest

»Schnelltest« is what the at-home COVID-19 test kit is called in Germany. I burned through an entire stack last week, only to find that the T-line became asymptotically thinner without ever vanishing completely. Thus, for the weekend, I decided to no longer check, which means I’m both positive and negative right now! I do feel a lot better, though, not counting ongoing headache attacks. Sometimes, it’s a migraine, and I have my triptan medication for that. Sometimes, it’s a both-sided COVID-19 headache, for which I have a well-stocked box from my last buying spree at a U.S. drug store. Sometimes, however, it’s something in between that just won’t go away, whatever I do—reminding me of the bad old days when triptan wasn’t available, or not available OTC. Days, fully wasted. But throughout my COVID-19 infection, the worst thing was the throat pain, which was so bad at times that even popping a coated Tylenol became the topic of the day.

Hence, not a lot of activity, but a bit more than the week before. At just drafts, I published my weekly AI/LLM/GPT Roundup on Lofty Ideals & Harsh Realities, followed by two topical linked-list items on, respectively, voice actors being forced to sign away their voice rights to AI and the future of AI-populated social media networks, aka Welcome to Hell. Finally, there’s a new album on Flickr with 25 shots from The Star Vista and neighboring structures in one-north/Queenstown, Singapore, and my daily vintage-style travel squaries on Pixelfed.

In case you haven’t already read about it on Mastodon, Twitter, Xing or LinkedIn, there will be the “Stochastic Parrots Day” event coming Friday, organized by The Distributed AI Research Institute, on the occasion of the second anniversary of the (in)famous “On the Dangers of Stochastic Parrots” paper. That paper got the co-leads of Google’s Ethical Artificial Intelligence team, Timnit Gebru and Margaret Mitchell, fired from Google. Together with Emily M. Bender and Angelina McMillan-Major, the paper’s other two co-authors, they will present and moderate discussions on AI hype vs. reality, data safety, worker exploitation, and other thrilling topics you rarely read about in AI hypetrain communiqués. The event takes place coming Friday at 16:00 CET (that’s Berlin time) or 3pm GMT/8am PDT; it’s online; and it’s free. Here’s the Eventbrite link to register, and here’s Timnit Gebru’s invitation post on LinkedIn. Looking forward to seeing y’all there on Friday!

Finally, the Sunday funnies, with a special homage to European office life: 5 surefire replies when someone addresses you in a meeting but you weren’t paying attention (German language); 10 things to say in an office kitchen (British English), and 13 phrases that mean nothing will happen (British English, too). Enjoy!

J.