Mastodon Feed: Posts

isagalaev ("Ivan Sagalaev :flag_wbw:") wrote:

RE: https://masto.ai/@phoronix/116476120066302561

OK, here's *the one* reason that'll finally make me find time and switch away from Ubuntu. I've been considering it for a while now, but was leaning towards taking the easy way out *again*: just upgrade my LTS and forget about it for 2 years. But "AI" is a bridge too far.

I don't car how non-envasive and opt-out it might be. One of the reason I use any Linux is to not be in adversarial relationship with my system.

Boosted by zkat@toot.cat ("Katerina Marchán"):
fastlydevs ("Fastly Devs") wrote:

Curious where WebAssembly is heading?

At Wasm I/O 2026, Luke Wagner shares the path to Component Model 1.0—and how you can start using it today.

🎥 Watch the talk: https://www.youtube.com/watch?v=qq0Auw01tH8

#WebAssembly #Wasm

baldur@toot.cafe ("Baldur Bjarnason") wrote:

Það er viðtal við Brynhildi Jenný (@uglyreykjavik.bsky.social) um myndasöguna sem hún gerði (og ég gaf út) á forsíðu RÚV. 😎

https://www.ruv.is/frettir/menning-og-daegurmal/2026-04-27-vantar-islensk-ord-i-myndasogur-467829

glyph ("Glyph") wrote:

By contrast, most of your audience will NOT have used an "AI" to detect a tumor, and hopefully those who *have* done so are sufficiently well-versed in the specific application that they use and the process for doing so that they are aware you can't just pop a scan into ChatGPT and go on with your day. If they jump into your mentions they're probably going to have some new and interesting information for you that can help inform your perspective.

glyph ("Glyph") wrote:

The flip side to this is that it's a good idea to avoid saying things that contradict your readers' direct experience. For example, the critic position that "AI" 'doesn't work'. Personally I kinda agree with this claim, given all the context and the evidence and sufficiently precise definitions of "work". But almost everyone have seen an "AI" app do something that looks like "working" and so saying this *without* all the shared context and definitions immediately discredits you.

glyph ("Glyph") wrote:

It feels like if you're entering a public debate, you have to perform a level of openness to be taken seriously. Someone who asserts their position while refusing to acknowledge the public consensus seems dogmatic, unresponsive to evidence.

But you can demonstrate this openness in other ways. "I'm open to evidence that 'AI' might be a net benefit to society, even given its enormous downsides, but right now, after so much evidence to the contrary that's an extraordinary claim"

glyph ("Glyph") wrote:

There is no need to begin your critical writing by saying "Well, I know AI has great uses, for example, in medical technology…" or "Surely it's going to change society…" or "It will take care of drudgery…" before your "but,"

If you're not an oncologist with experience using "AI" to spot tumors, you don't actually *know* that it has great uses for that. Nobody knows if it's going to "change society", so you don't have to react to baseless claims from "futurists". You can just… not.

Boosted by cstanhope@social.coop ("Your weary 'net denizen"):
ThompsonArt@mastodon.art ("Aled Thompson") wrote:

Green European Woodpecker

Prints available here -
https://www.inprnt.com/gallery/thompson%5Fart/european-green-woodpecker-2025/

Boosted by cstanhope@social.coop ("Your weary 'net denizen"):
brianbilston@mastodon.online ("Brian Bilston") wrote:

Today’s poem is called ‘Why You Should Never Ever Follow Your Dreams’.

glyph ("Glyph") wrote:

There is a human tendency to grant our interlocutors empathy by default. This can be weaponized by bad-faith actors trying to move the overton window.

Someone says "AI can do X" and it's uncomfortable to say "no, I already know it can't" or even "prove it". You hear it over and over again from a hundred "journalists" acting as stenographers for big tech, and you feel like you have to grant that position grace.

You don't have to fall for it *AND* you don't have to be mean about it.

db@social.lol ("David Bushell 🪿") wrote:

"GitHub AI Credits*" 💀

https://github.blog/news-insights/company-news/github-copilot-is-moving-to-usage-based-billing/

* redeemable at participating locations only, no cash value, ToCs apply

cstanhope@social.coop ("Your weary 'net denizen") wrote:

(It was not one of my boards. Phew!)

cstanhope@social.coop ("Your weary 'net denizen") wrote:

It's always a little unsettling to walk into work on a Monday morning and be greeted with the odor of burning electronics. "I sure hope that's not one of my boards!"

Boosted by glyph ("Glyph"):
dreid@wandering.shop wrote:

RE: https://masto.ai/@phoronix/116476120066302561

Canonical's ability to do something stupid that will annoy their users and also being late to the hype party is remarkable.

cstanhope@social.coop ("Your weary 'net denizen") wrote:

I think for once we should put aside putting aside the ethical concerns...

Boosted by cstanhope@social.coop ("Your weary 'net denizen"):
falseknees ("False Knees") wrote:

Intelligent

Boosted by adele@social.pollux.casa ("Adële 🐁!"):
vitex@f.cz ("Vitex") wrote:

@adele

#Ubuntu #Debian #packages ready to download and install

https://repo.vitexsoftware.com/pool/main/s/smolfedi/

zkat@toot.cat ("Katerina Marchán") wrote:

my migration from this instance to @zkat@fedi.zkat.tech seems to have frozen, and my followers have stopped migrating. Anyone know how to get this unstuck? The latter is a GTS instance.

Boosts welcome!

dysfun@treehouse.systems ("gaytabase") wrote:

The era of proof abundance

oh god, it is absolutely cringe hearing supposed legends talk this way about bullshit generators

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
grimalkina ("Cat Hicks") wrote:

And while you're waiting for the book, here's my newsletter: https://www.fightforthehuman.com/

Boosted by baldur@toot.cafe ("Baldur Bjarnason"):
grimalkina ("Cat Hicks") wrote:

And was just reminded to post this, lol, here's my book!!! The Psychology of Software Teams, it's coming out in July, you can already pre-order it at some online retailers, and it has a WHOLE CHAPTER on "The Performance Paradox" to help you become a *resilient* high performer :)

https://www.routledge.com/The-Psychology-of-Software-Teams/Hicks/p/book/9781032963389

jscalzi@threads.net ("John Scalzi") wrote:

I have been fortunate that even before I was a full-time novelist I was making a good living doing freelance and consulting for companies on writing/editing matters. My whole professional life has been as a writer. What helped (and still does) was that I was not precious about what I would write. As long as it did not run against my personal ethics, and you would pay me my rate, I would write it for you. That was basically my first 20 years as a writer.

RE: https://www.threads.com/@spirit.horse.tarot/post/DXo28L%5FFDMQ

Boosted by jsonstein@masto.deoan.org ("Jeff Sonstein"):
wdlindsy@toad.social ("William Lindsey :toad:") wrote:

"Could we please take a moment to remember that the largest, bloodiest, violent attack on any Washington institution in modern memory was actually led by…Trump. On January 6, 2021, he incited an assault on the Capitol that resulted in multiple deaths, scores of injuries and even a call for assassination of Trump’s own Vice President."

~ David Rothkopf

#WHCD #Trump #media #violence #Jan6 #insurrection
/5

https://davidrothkopf.substack.com/p/five-uncomfortable-truths-about-the

Boosted by glyph ("Glyph"):
grimalkina ("Cat Hicks") wrote:

I have said and will continue to say most people have zero idea what it's like for scientists in the US right now.

We sat at an outdoor table the other day with a brilliant older queer scientist friend and ran into several scientist friends including one who worked at govt agencies on science funding. It is like having conversations after an apocalypse. So many people lost, labs folded, international postdocs gone, lines of work canned. DEI work going undercover, forbidden words erased

Gargron ("Eugen Rochko") wrote:

Male fragrance sections at shops are always so sad. Would you like to smell like you just downed a bottle of gin, like a leather couch, or like you were sitting on the wrong side of a campfire? Women get all the nice fruits and flowers. This stuff shouldn’t be gendered anyway like most things.

Boosted by dysfun@treehouse.systems ("gaytabase"):
partim@social.tchncs.de ("Martin Hoffmann") wrote:

And another court found Germany’s border checks to be unlawlful.

A law professor was checked when returning home from the conference “40 years Schengen Treaty” and, presumably overwhelmed by the irony of it all, sued.

This is important since this is the first case after the reform of the Schengen border codex in 2024 – and the court still found the reasons for reintroducing border checks insufficient.

This was a decision by a lower court, will likely be appealed, and is expected to eventually wind up at the European Court of Justice.

Until then, the federal government sees no reason to stop its unlawful behaviour.

German news report (thanks, @mikey179): https://www.swr.de/swraktuell/rheinland-pfalz/urteil-grenzkontrollen-zwischen-deutschland-und-luxemburg-rechtswidrig-100.html, full decision: https://vgko.justiz.rlp.de/fileadmin/justiz/Gerichte/Fachgerichte/Verwaltungsgerichte/Koblenz/Entscheidungen/Nr%5F10-2026%5FVOE%5F3%5FK%5F650-25%5FKO%5FUrteil%5Fvom%5F27.04.2026.pdf

pzmyers@freethought.online ("pzmyers 🕷") wrote:

The new American university model of education.

https://freethoughtblogs.com/pharyngula/2026/04/27/a-stifling-ignorance/

dysfun@treehouse.systems ("gaytabase") wrote:

i am re-reviewing graphing libraries in javascript and so far i've got

• huge
• crap
• by a company advertising ai bullshit

Boosted by slightlyoff@toot.cafe ("Alex Russell"):
bkardell@toot.cafe wrote:

Web Serial stuff experimental in Firefox 🤯

https://developer.mozilla.org/en-US/docs/Web/API/SerialPort/writable

dysfun@treehouse.systems ("gaytabase") wrote:

i did also get fairly far with making avx-512 insert routines for the HOT when i last touched it. i think it might even be possible to do the core of the insert and delete procedures in constant time for each node layout. the original insert procedure from the HOT paper is a branchy, loopy mess (and lo and behold, it advertises itself for read-mostly workloads!), but fortunately it seems largely amenable to simdification if you've got avx512 (and thus predication) available.

if everyone had avx-512, the code required to implement the HOT in C would probably be quite small. you might not even hate to maintain it. but already you're going to need a portable branchy loopy mess, so that's two versions. throw in support for a few more cpus, each with harshly increasing restrictions on what you can do and it becomes a mess quite quickly.