Ne, das hat @Korne127 schon auf den Punkt gebracht. Menschys die Causa X unterstützen kritisieren, weil nicht genug gegen Causa Y getan wird ist kein Gewinnerzug.
Intersektionalität du lernen musst, um eine geschlossene Front gegen die Reaktion zu zeigen.
Um originalen Pfostierer mal auf eine innovative Art zu dissen:
arbeitete an etwas, das sich nicht einfach speichern lässt
Welch primitive Technologie ist das denn? Alle ernstzunehmende Rechenknecht-Arbeit sollte in Textdateien speicherbar sein, um bei Bedarf im Versionskontrollsystem abgelegt zu werden!
Cooler Käfer! In anderen Neuigkeiten, mein Vertrauen in unsere gesamte Rechenknecht-Infrastruktur hat weiter gelitten.
whatever minor body of irrelevant five-follower internet loons might bother trying to argue the literal uselessness of a predictive text generator
/c/techtakes mention!?
Wenn man bewusst einen Überblick über aktuelle Geschehnisse bekommen will, ist ein Besuch auf https://dw.com/ (weltweit) und https://tagesschau.de/ (🇩🇪) immer zu empfehlen.
Nein, das ist ganz falsch! Studien haben belegt, dass in den 5 Sekunden zwischen dem Öffnen der Schachtel und dem ersten Bissen 95% der Vitamine durch Oxidation zerfallen. Nur wenn man in einem Bissen den Burger samt knackiger Schachtel abbeißt, erhält man den vollen Nährwert.
Ihr könnt mir da vertrauen, meine Cousine ist Burgerologin!
Zum Glück hab ich den pfostierten Text nicht gelesen, und konnte dieses erquickende Christoph Waltz MichMich unironisch genießen.
Ob sogenannte soßen-offene GenKI wirklich soßen-offen ist, ist diskutabel.
zlng
Modelle mit offenen Gewichten sind eher mit Freewarezu vergleichen, und auch wenn Trainings-Algorithmen offen sind, können sie nicht ohne BigTech Infrastruktur und Datensätze rekreiert werden.
Davon unabhängig bin ich der Meinung, dass GenKI Nutzung die Fähigkeit zum kreativen Ausdruck verkümmern lässt, hervorragendend visualisiert vom Nutzenden @bobtimus_prime@feddit.org .
Yeah, the same thing struck me. I'd guess they were jumping on the buzzword, but x-risk just was deemed gaudy and unserious.
Might be semi-related: the german aerospace/automotive/industrial research agency has an "AI Safety" institute (institute = top level department).
I got a rough impression from their website. They don't seem to be doing anything that successful. Mostly fighting the unwinnable battles of putting AI in everything without sucking and twiddling machine learning models to make them resilient against malicous data. Besides trying to keep the torch of self-driving cars alive for the german car industry. Oh, and they're doing the quantum AI bit.
They're a fairly new institute, and I heard rumors they're not doing great. Maybe the organization resists the necessary insanity to generate new AI FOMO at this point. One can dream.
Der Antisemitismusbeauftragte dea Bundes, Felix Klein, findet harte Kritik an Israel und Fragen nach dem Völkerrecht überprüft Notizen nicht antisemitisch!
I was going to shitpost that Trump is the least neo-colonial president cuz he cut all foreign aid, but I realized I kinda believe that unironically. I'm in the anti-death-and-suffering camp of course, but a hundred kinda self-serving national aid programs might just not cut it.
(This might be inspired by the Merz government planning to roll the special development aid office into the foreign affairs ministry, partly to tie it more strongly to national interest.)
Maybe we need to bring back the UN bigly.
A few months ago, Casey Newton of Platformer ran a piece called "The phony comforts of AI skepticism," framing those who would criticize generative AI as "having fun," damning them as "hyper-fixated on the things [AI] can't do." I am not going to focus too hard on this blog, in
I haven't watched it. I don't know how well she will cover the subject or how deep the rabbit hole she will venture.
All I know is she's delightful and I sure as hell won't read that bilge myself, so I'm looking forward to an entertaining summary.
Edit: I watched it. I had a good time.
Soft-serve machines get a not-quite-parfait exemption to DMCA circumvention rule.
Character.AI Teen Suicide Lawsuit refers to a controversy and lawsuit against the user-generated artificial intelligence (AI) chatbot website Character.AI
archive of the mentioned NYT article
Businesses using artificial intelligence to generate code are experiencing downtime and security issues, according to Sonar CEO.
cross-posted from: https://sh.itjust.works/post/25239919
So I recently got an excuse rant about my opinions on federated tech. I think it's pretty much the best we can hope for in terms of liberating tech, with very few niches where fully distributed tech is preferable.
Needing a server places users under the power of the server administrator. Why do we bother? "No gods, no masters, no admins!' I hear you shout. Well, there's a couple reasons...
Maybe using software is just an intrinsically centralized activity. One or a few people design and code it, and an unlimited number of people can digitally replicate and use it. Sure, it may be free software that everyone can inspect and modify... but how many people will really bother? (Nevermind that most people don't even have the skills necessary.)
Okay, so we always kind of rely on a central-ish dev team when we use tech. Why rely on admins on top of that? I believe the vast vast majority of people doesn't have the skills and time to operate a truly independent node of a fully distributed tech. Let's take Jami as an example:
"With the default name server (ns.jami.net), the usernames are registered on an Ethereum blockchain."
So a feature of Jami is (for most users) implemented as a centralized service. Yikes. You could build and run your own name server (with less embarrassing tech choices hopefully), but who will really bother?
But say you bothered, wouldn't it be nice if your friends could use that name server too, and gain a little independence? That sounds a lot like decentralized/federated tech.
Keeping a decent service online is a pain in the butt. Installing SW updates, managing backups, paying for hardware and name services... nevermind just the general bothering to understand all that mess. And moderation, don't forget moderation. I'm saying it's not for everyone (and we should appreciate the fuck out of [local admin]).
I believe that servers and admins are our best bet for actual non-centralized tech. A tech-literate person tending a service for a small- to medium-size community is much more feasible than every person running their independent node (which will probably still depend on something centralized).
And maybe that's just the way we bring good ol' division of labour to the Internet. You have your shoemaker, your baker, your social media admin. A respectable and useful position in society. And they lived happily ever after.
Questioning the design and dev progress was apparently "against company policy."
Apparently a senior SW engineer got fired for questioning readiness of the product, dude must still be chuckling to himself.
Found the story here https://hachyderm.io/@wesley83/112572728237770554