lvxferre

joined 2 years ago
MODERATOR OF
[–] lvxferre@mander.xyz 7 points 1 day ago

Fuck neoliberalism. I might not like those kids giving thumbs up to authoritarianism but they're less worse than neolib, it's like comparing elephant shit with cat shit.

That said: a few Hexbear users are fairly reasonable, but plenty of them remind me 11yos. Specially in groups. db0 threw the bait, in a humour community for anarchists... and they all ate the bait, with malagueta sauce. No, wait, Scorpion Trinidad sauce. While screaming "I CAN STAND THE HEAT!" amidst their torrential tears.

Eventually the thing gets old, db0 gets bored and tells them to disengage.

It was funny that the nanny - who was supposed to take care of the kids - was there too. And he's another 11yo! ~~He forgot to edit his comment and say EDIT WOW THANKS FOR THE GOLD KIND STRANGER! They might've left Reddit, but they still behave more like redditors than like decent people. Also, I saw a neoliberal Anglo supremacist there. Full of gullibleness. He still didn't think on the purpose of this utterance. And yes, I'm still laughing my arse off that muppet.~~

I'm almost sure db0 knows why those kids are mostly defederated. And the reason his instance doesn't do it is because, good or bad, they're left-wing presence in the fediverse.

Hexbear users who might take offence with my comment are encouraged to address their complains about their precious feelings being violated to this form.

[–] lvxferre@mander.xyz 2 points 1 day ago

It's basically my experience with translation, too: asking a LLM is a decent way to look for potential ways to translate a specific problematic word, so you can look them up in a dic and see which one is the best. It's also a decent way to generate simple conjugation/declension tables. But once you tell it to translate any chunk of meaningful text, there's a high chance it'll shit itself, and output something semantically, pragmatically, and stylistically bad.

[–] lvxferre@mander.xyz 16 points 1 day ago (1 children)

I'm not. You can't lose trust on something if you never trusted it to begin with.

I. Talent churn reveals short AGI timelines are wish, not belief

Trying to build AGI out of LLMs and similar is like trying to build a house, by randomly throwing bricks. No cement, no foundation, just the bricks. You might want to get some interesting formation of bricks, sure. But you won't get a house.

And yes, of course they're bullshitting with all this "AGI IS COMING!". Odds are the people in charge of those companies know the above. But lying for your own benefit, when you know the truth, is called "marketing".

II. The focus on addictive products shows their moral compass is off

"They", who? Chatbots are amoral, period. Babbling about their moral alignment is like saying your hammer or chainsaw is morally bad or good. It's a tool dammit, treat it as such.

And when it comes to the businesses, their moral alignment is a simple "money good, anything between money and us is bad".

III. The economic engine keeping the industry alive is unsustainable

Pretty much.

Do I worry that the AI industry is a quasi-monopoly? No, I don’t understand what that means.

A quasi-monopoly, in a nutshell, is when a single entity or group of entities have an unreasonably large control over a certain industry/market, even if not being an "ackshyual" monopoly yet.

A funny trait of the fake free-market capitalist that O’Reilly warns us about is that their values are always very elevated and pure, but only hold until the next funding round.

That's capitalism. "I luuuv freerum!" until it gets in the way of the money.

IV. They don’t know how to solve the hard problems of LLMs

Large language models (LLMs) still hallucinate. Over time, instead of treating this problem as the pain point it is, the industry has shifted to “in a way, hallucinations are a feature, you know?”

Or rather, they shifted the bullshit. They already knew it was an insolvable problem...

...because hallucinations are simply part of the LLM doing what it's supposed to do. It doesn't understand what it's outputting; it doesn't know if glue is a valid thing to add to a pizza, or if humans should eat rocks. It's simply generating text based on the corpus fed into it, plus some weighting.

V. Their public messaging is chaotic and borders on manipulative

O rly.

Stopped reading here. It's stating the obvious, and still missing the point.

[–] lvxferre@mander.xyz 4 points 1 day ago

And people who don't have bread to eat should eat cake instead. *sigh*

[–] lvxferre@mander.xyz 16 points 1 day ago* (last edited 1 day ago) (2 children)

The website's quality decline has been for ten years or so. I remember well how it used to be; if I had my first account there it would be, like, 16 years old? (Good riddance.)

I'm glad people mentioned PieFed there (+10 score as of now!), as well as old style forums. (I'm hoping those eventually join the Fediverse. But even if they don't, they still have a place in my heart.)

[–] lvxferre@mander.xyz 2 points 2 days ago

After Denmark and Schleswig-Holstein, now it's Lyon.

I expect "[insert European government] ditches Microsoft" to become more and more common news, until it becomes non-noteworthy.

[–] lvxferre@mander.xyz 4 points 2 days ago

Group name: None, individual drawing. I'm just reserving the space in case anyone wants to do stuff that interacts with it (like critters climbing it, etc.)
Matrix: none
Template: link
Fun fact: it's a tree. Very wood. Much leafy.

[–] lvxferre@mander.xyz 0 points 2 days ago* (last edited 2 days ago)
  1. Group name: Parabellum.
  2. Matrix: https://matrix.to/#/#parabellum-canvas:matrix.org
  3. template URL: N/A, see note.
  4. Fun fact: named after Caesar's "si uis pacem, para bellum" (if you want peace, prepare for war).

NOTE: the group's goal is to coordinate attacks against the largest country flag in the canvas, whichever it is. Void, random doodling, etc. is welcome.

further info, including channel rules.Channel rules:

  1. Be a reasonable and decent human being. (e.g. no bigotry)
  2. No divisive off-topic or NSFW content.
  3. No fragmenting the movement from the inside: apology to large country flags, JAQing off, sealioning, etc. This will be strictly enforced.
  4. No cheating, doxxing, or encouragement of.
  5. Have fun.

The target is always the biggest country flag. Drawings in the flag are disregarded; only width x height counts. In the case of a tie, the leftmost flag takes precedence. Focus preferably on most iconic features of the flag.

Observers are welcome, as long as they follow the rules. You're also free to join the group even if you belong to another group, just make sure there's no conflict of interest.

Q-and-AQ1: Why would I join it?
A: Reasons are up to the individual. We don't need to share the same motivation, only the goal.

Q2: What constitutes a large or small flag?
A: Up to debate but we don't need to set this line up, since we're picking on the largest one anyway.

Q3: What about the trans flag? Or the anarchist one?
A: Not country flags, we're leaving them alone.

Q4: What about drawings that are not part of the flag, but associated with it?
A: Leave them alone, and focus on the flag itself. And in the case the flag is built into a bigger drawing, focus on the elements of the flag, not of the rest of the drawing.

Q5: Vandalism is wrong!
A: Not a question, but whatever. This sort of online canvas is not just cooperation, but also competition. Also, people actually concerned about vandalism in the canvas know that the main source of said vandalism are large country flags.

Q6: If u dun liek it, than y u dun fuck off teh gaem?
A: No. People who want to get rid of large flags have as much right to organise themselves towards their common goal as you have to organise your flag. Deal with it.

Q7: as someone working on a large flag, I ask u, can u stop doin that?
Q8: excuuuuuse me, do you know that I'm entitled to have you listen to my defence of large flags?
Q9: ackshyually is this harassment?
No.

Q10: why do you hate [insert population]?
A: Our target is decided based on canvas area, not personal feelings towards the subjects of the government associated with that flag.

Q11: instead of fighting us, why don't you join us to build our MONUMENT?
A: I do monuments like this every day. In my bathroom, just after #2.

[–] lvxferre@mander.xyz 5 points 2 days ago

Yes. All my yes.

[–] lvxferre@mander.xyz 17 points 2 days ago (2 children)

I think that, for most dbzer0 users (and quite a few landlubbers like me), being able to choose between a Lemmy and a PieFed instances is only positive. What you need to take into account is the additional work this incurs towards you guys, the admins; you'll be maintaining two instances instead of just one.

[–] lvxferre@mander.xyz 11 points 2 days ago* (last edited 2 days ago) (1 children)

If my analysis is correct, the people in the Ukrainian government do know that Putin and Trump are on the same team. And they're trying to break it, by exploiting the fact Trump has a huge ego and a small brain. That's why Podolyak is framing the attacks as humiliating Trump.

After Zelenskyy was humiliated in the White House no less, he must know these people are Demons right?

When fighting a devil, sometimes you need to make a deal with another.

[–] lvxferre@mander.xyz 8 points 2 days ago

I'm low-key excited. It has been a long time since the anime series, it was a blast, and based on this trailer it seems the movie keeps rather well the "everything is so fucked up might as well laugh" feel.

 

EDIT: @mindbleach@sh.itjust.works shared something that might help to circumvent this shit:

Contained in these parentheses is a zero-width joiner: (​)

Basically, add those to whatever you feel that might be filtered out, then remove the parentheses. The content inside the parentheses is invisible, but it screws with regex rules.

 

Changes highlighted in italics:

  1. Instance rules apply.
  2. [New] Be reasonable, constructive, and conductive to discussion.
  3. [Updated] Stay on-topic, specially for more divisive subjects. Avoid unnecessarily mentioning topics and individuals prone to derail the discussion.
  4. [Updated] Post sources whenever reasonable to do so. And when sharing links to paywalled content, provide either a short summary of the content or a freely accessible archive link.
  5. Avoid crack theories and pseudoscientific claims.
  6. Have fun!

What I'm looking for is constructive criticism for those rules. In special for the updated rule #3.

Thank you!

EDIT: feedback seems overwhelmingly positive, so I'm implementing the changes now. Feel free to use this thread for any sort of metadiscussion you want. Thank you all for the feedback!

 

Apparently humpback whale songs show a few features in common with human language; such as being culturally transmitted through social interactions between whales.

"The authors found that whale song showed the same key statistical properties present in all known human languages" - my guess is that the author talks about Zipf's Law, that applies to both phoneme frequency and word frequency in human languages.

[Dr. Garland] "Whale song is not a language; it lacks semantic meaning. It may be more reminiscent of human music, which also has this statistical structure, but lacks the expressive meaning found in language." - so while it is not language yet it's considerably closer to language than we'd expect, specially from non-primates.

 
 

Based on

SVG source for anyone willing to give it a try. Made with Inkscape. The emojis were added as images because Inkscape.

 

It's a 10m papyrus scroll from Herculaneum, one of the cities buried by Vesuvius' volcanic ash in 79 CE. It's fully carbonised but they're using a synchrotron to create a 3D model of the scroll without damaging it. Then they're using AI (pattern recognition AI, perhaps?) to detect signs of ink, so they can reconstruct the text itself.

The project lead Stephen Parson claims that they're confident that they "will be able to read pretty much the whole scroll in its entirety". And so far it seems to be a work of philosophy.

 

The title is a bit clickbaity but the article is interesting. Quick summary:

A new ancient population was recognised, based on genetic data. This population has been called the Caucasus-Lower Volga population, or "CLV". They were from 4500~3500BCE, tech-wise from the Copper Age, and lived in the steppes between the North Caucasus and the Lower Volga. .

About 80% of the Yamnaya population comes from those people; and at least 10% of the ancestry of Bronze Age central Anatolians, where Hittite was spoken, also comes from the CLV population. The hypothesis being raised is that the CLV population was composed of Early Proto-Indo-European speakers (the text calls it "Indo-Anatolian").

103
submitted 5 months ago* (last edited 5 months ago) by lvxferre@mander.xyz to c/growthefediverse@slrpnk.net
 

I'm sharing this pic because it might be useful, to advertise Lemmy in Reddit meme communities and the likes. It isn't supposed to be a full info dump, just to spread the word that Lemmy exists and give people some room to ask questions about it.

The copypasta is from @Blaze@lemmy.dbzer0.com. The meme is from @JokkaJukka@lemmy.world.

Here's the source SVG file in case anyone wants to edit it.


EDIT - @Libb@jlai.lu had a great take on this idea, I need to share it here:

 
 

cross-posted from: https://quokk.au/post/1499265

What a Christmas present!

Italo-Celtic is a hypothetical branch of the Indo-European languages. If that branch is real, it means that the Italic and Celtic languages are closer to each other than to other Indo-European languages.

This hypothesis has been raised multiple times in the past, due to a few shared morphological features between Italic and Celtic languages; for example, the *-ism̥mo- superlative. But that's on its own weak evidence, so this genomic data makes wonders to reinforce this hypothesis.

And also to bury the competing (IMO rather silly) Italo-Germanic one.

Graeco-Armenian is similar to the above, but between the Hellenic languages and Armenian. There were lots of competing hypotheses "tying" both branches to other "random" Indo-European branches; for example I've seen Indo-Greek, Italo-Greek, Armeno-Germanic, Armeno-Albanian...

 

In case anyone wonders about the star:

what you type what you get
=
* 💢
$
<
>
@
 

1.6 introduced two amazing features to make mad money from fishing.

Pic related - it's the first Summer, I didn't ship all the fish (I couldn't smoke all of them in time), and I'm making 20kg/day, even being bad at fishing.

The bait maker is unlocked by Fishing Lv6. It requires 3 iron bars, 3 coral, 1 sea urchin. You put a fish there, and get 5~10 pieces of bait targetting that fish.

But why would you waste fish to get more fish, instead of using bug meat for the bait? Because you can target the most lucrative fish - like I did with sturgeon there, if I used normal bait I'd be getting carps instead.

Then there's the fish smoker. The recipe is from Willy's and it costs 10kg. You need 10 hardwood, 1 river jelly, 1 sea jelly, 1 cave jelly.

The river and sea jellies are not a big deal to get. The hardwood will likely require luck in the mines or an upgraded axe. Cave jelly is a pain, you need to fish in the mines for a while. (Ideally at Lv100, but Lv20 does the trick.)

But oh wow. It's worth. It doubles the price of the fish, but it keeps quality. Check the iridium quality sturgeon, it's 1200g! Lucrative fish tend to be unruly, but if you fish at good spots you can reliably get gold quality fish; 1.5 from gold quality * 1.5 from angler profession * 2.0 from smoking = 4.5 times the original price.

It does require coal but once you got the cash rolling you can buy it from Clint, that's what I'm doing.

view more: ‹ prev next ›