Notices by Dr. Juande Santander-Vela (juandesant@astrodon.social)

@rgulick @bmacDonald94 learn it in 2016, happy to see that the effort pays out a second time…

ChatGPT does not think like us. They are using anthropomorphic designations (like calling it conversations, chats, or thinking) for additional constraints on Large Language Models (LLMs) based on Transformer-architecture. That is not the way humans think.

Please stop the anthropomorphic treatment of LLMs.
https://flipboard.com/@fastcompany/fast-company-l36v8lfpz/-/a-Usz5CNOqQ1SufSL6C6KB3w%3Aa%3A3199563-%2F0

Striking! Direct radio imaging of a star not in our direct vicinity showing how it evolves with time… the animation is not computer generated, is the result of this direct imaging!

More information and a more detailed video (including dates for the observations) can be accessed through the link below:
https://www.almaobservatory.org/en/press-releases/astronomers-track-bubbles-on-a-stars-surface-in-the-most-detailed-video-yet/

#ALMA #AtacamaLargeMillimeterSubmillimeterArray #AtacamaLargeMillimeterArray #AstroSeismology #StarEvolution #RDoradus

@thomasfuchs how much time have you integrated for it?

Thanks for the link, @jbzfn, but I think it is important to emphasize the main message from the original article, not the summary from @phoronix: the *BSDs have their own fork of with a richer feature set, and better generic acceleration compatibility.

The question is about when does BSD stop supporting that fork, and for now the answer is that they will still do it…

Original article in the NetBSD blog:
https://blog.netbsd.org/tnf/entry/x_org_on_netbsd_the

@thomasfuchs no offense to @matt, but that Meta Quest looks to me like Bane’s muzzle, just on the wrong part of the head…

@matt well, I think there are a couple subtle differences between this picture and the one from Tim Cook on Vanity Fair… hope @thomasfuchs sees the difference ;-)

@mac84tv is that a CHRP machine? It does not look bad at all…

Happy new year of the One Dot Leader (aka "․", aka U+2024)

🎉🎆🎇2️⃣0️⃣2️⃣4️⃣

Today I learned about “schizmogenesis”, the fancy name for perceiving something your opponent’s likes/dislikes as something you have to dislike/like.

Very cogent article from @pluralistic.

https://pluralistic.net/2023/11/11/ad-jacency/#brand-safety

#TIL #schizmogenesis

@feditips @fdroidorg you can also do that in @ivory and @tootapp, at least.

@feditips @SmartmanApps @jherazob I think it takes more time now… here is how I see it in the web interface…

@SmartmanApps @feditips @jherazob you can change the link to a yewtu.be link.

@thomasfuchs sorry, I think we’re both in agreement. It is irresponsible not to allow already established public emergency services, both morally and financially, because it removes incentives to use the birdsite.

But it is also wild that existing infrastructure with public access is not use more. You can geofence alerts, and even if you silence them (it is different from the volume and mute switches), hearing everyone else’s phone beeping and vibrating with a particular sound is telling.

@thomasfuchs this still begs the question: why don’t the use the same underlying mechanism of the AMBER alerts? That is what the disaster prevention and mitigation department (SERNAPRED, formerly ONEMI) does in Chile… and the mechanism is available internationally.

@thomasfuchs why, why???

@evan defining “able to program” as both “not being restrained by lack of tools at different expertise levels” and “there are ways for people to build up their expertise level”. Not as “everyone should be required to program”.