Leo Betschart @betschart@qoto.org

This is an experiment. Please boost.

http://mastoddosnt.ddns.net/

Here's the idea: This post is going first to my followers, then, if they boost it, to other people. This domain has been registered for only this experiment. I should see in my web server's logs when mastodon instances start crawling the site for info. Then maybe also some curious humans.

I just want to play with my monitoring a bit :)

Good morning #medlibs, friendly reminder of how it works when a subject heading can have narrower terms or not have NTs depending on which hierarchy branch you are referencing

3,500 year old cheese - likely oldest ever found - on necklaces in mummies! Analysis in a report in Cell

"World’s oldest preserved cheese found in necklaces on mummies in China"
via @popsci

https://www.popsci.com/science/worlds-old-cheese-china-mummies/

#Cheese #Science #Dairy #FoodHistory

I want to remind y'all the only reason I can legally tweet out discovery evidence is that I repeatedly refused to sign the NDA Apple tried to force on me, & filed an NLRB charge against Apple for still claiming labor disputes are Apple Confidential.

Apple only released the handful of material docs I have now after I filed NLRB charges about their discovery shenanigans. They continue to pressure me to sign a protective order. I keep telling them their public safety hazards are not confidential.

Announcing Zotero 7, the biggest update in Zotero’s 18-year history

https://www.zotero.org/blog/zotero-7/

Reading a paper in Information Matters

"Being an expert in information should surely facilitate professional information practices."
You might think that, until you see a library's shared drive ;)

The whole AI thing has me endlessly confused. Half the market is crashing because investors didn't see any signs of payoff in the quarterly earnings report, but I'm so lost as to what exactly they were expecting to see. Did they just not pay any attention at all to what these companies were actually doing with AI?

Were they expecting exponential Instagram usage growth as a result of Meta making it so you can have a conversation with the search bar? Or maybe everyone was going to buy 10 new Windows licenses in celebration of Microsoft announcing they want to install AI powered spyware on everyone's computer? Or was Google going to sell more ads by replacing all the search results with reddit shitposts. Either I'm missing something or everyone's 2 remaining brain cells are just really busy fighting to death for 3rd place.

@SRLevine I hate plugging my own stuff…but I would recommend looking at some of the magic solvent pairs with DMSO as an additive or co-solvent https://www.sciencedirect.com/org/science/article/abs/pii/S108361602101834X

Being a reader in 2024 is a really lonely place. I have read several absolutely incredible novels this year and trying to get my friends to just try one is like pulling teeth. Many of these stories are about places and activities close to us, too. They aren't long. They aren't tortured reads... and yet.

It's so interesting how when you're a female leader you're never technical enough until you are and then you're too rigorous and inflexible. Love this labyrinth with no exit for us.

@grimalkina I feel this so hard.

"Be assertive, but not like that"

"Be confident! Oh, sorry, that's too confident"

"Own your expertise! Wait that's too many areas of expertise"

"Get things done! But only via collaboration with people who don't want to collaborate with you"

"You can always speak candidly and transparently here, but we expect high quality communication, which is why we'll interpret everything you say in the most adversarial way possible"

Sigh

Les toilettes inclusives c’est aussi pour la noblesse française.

#MarieAntoinette

Many speak idyllically about a world with "fair" or "unbiased" LLMs, but is that even possible? In our new preprint, we take the most well-defined principle of AI safety/ethics and show, in reality, an LLM could never be fair under any definition in the current ML literature. [Brief thread:👇; paper link: https://arxiv.org/abs/2406.03198]

@kravietz
As you point out it can be difficult on administrative grounds for organizations or businesses to make a donation outright. One approach that worked for me as a (now retired) manager, and I mention it for readers who may be in a position to act, is to identify needed features or fixes that are relevant for your organization and finance that work. Then you have a business case and measurable deliverables for the bean counters.

Few things piss me off more than a huge, multi-billion IT corporation that suddenly sends me an email regarding an open-source project I’ve been running since 1990’s that I’ve recently shut down due to absolute lack of interest from its users… which happened to be telcos and large IT companies. Here’s what I replied:

Thank you for your email. As it’s often the case with open-source projects, their value to organisations is only noticed and appreciated when they go offline. I have maintained pam_tacplus for the last years and it had the call for sponsorship prominently displayed for most of the time specifically because it’s a legacy project that is difficult to maintain. None of the commercial companies that clearly do rely on it ever demonstrated any interest in even nominal donations, so it was archived. While it’s notable someone finally noticed it, I’m not the person to discuss any future development any more.

I did work in large companies and I do understand the sick logic that drives them, when it’s easier to get approval for annual spending of $50k for some office decorations than $100 for a mission-critical project which happens to be open-source and can be used for free for some time.

But it’s possible. If you’re working in such roles, please make every effort to get this $100 because otherwise it will become your responsibility to develop and maintain code that you always got for free.