Mechanize

@Mechanize@feddit.it

I did nothing and I’m all out of ideas!

This profile is from a federated server and may be incomplete. Browse more on the original instance.

Mechanize,

Yeah, but don’t forget that they poked him with a sword! A special sword!

That must mean something!

French MPs want to amend EU's copyright rules to cover generative AI (www.euractiv.com)

"How to identify original works by artists? How to attribute works generated by AI intermediaries? How to remunerate authors whose works have been used? How to manage opt-outs for artists who refuse their content to be used by AI? These are the questions that require a review of the copyright directive in light of generative...

Mechanize,

I like the idea of 24 years plus the possibility of an extension with a fee, or proof of activity, for another 12 years if the owner is a person.

If, at any point, the rights are sold, passed, or the owner is not a person (but for example a corporation or association) it should last 12 years from that time with an extension of 6, so the ability to sell your idea or give it to your spawns for a soft landing is not destroyed, but it can not be abused.

To further avoid abuse if it has already been extended before being sold it should last only 6 years without possibility of extension. So at most it would be 42 years for active and valuable things.

EDIT: To clarify, if it gets sold multiple times, the “timer” shouldn’t be reset but keep ticking down from the date of the first transaction, or it would open the door to accounting shenanigans just to keep it alive and locked.

Mechanize,

…is that a Goa’uld? Scary!

Mechanize,

Yeah. GDPR should have been implemented as a mandatory part of HTML or even HTTP that interacts with a builtin browser feature.

Well, it kind of is. The Do Not Track header has recently seen a court win in Germany (source):

It turned out that the judge agreed with vzbv, ruling that the social media giant is no longer allowed to warn users it doesn’t respect DNT signals. That’s because, under GDPR, the right to opt out of web tracking and data collection can also be exercised using automated procedures.

And it is basically the same in California too Source

GPC is a valid do-not-sell-my-personal-information signal according to the California Consumer Privacy Act (CCPA), which stipulates that websites are legally required to respect a signal sent by users who want to opt-out of having their personal data sold.

informapirata, (edited ) to main Italian

Una riflessione sui successi e gli insuccessi del progetto feddit.it e le prospettive del lemmyverso e del bloggingverso. E voi che ne pensate?

https://www.informapirata.it/2024/01/14/feddit-it-tutte-le-criticita-e-le-potenzialita-dellalternativa-italiana-a-reddit-tra-bloggingverso-e-fediverso/

Mechanize,

feddit.it non è cresciuto tanto quanto le altre istanze perché, …, è cresciuto meno delle altre istanze.

E questo è tutto, vostro onore.

Scherzi a parte è vero che - in generale - il fediverso è piccolo, mal indicizzabile e la regola dell’1% è abbastanza brutale, ma ad esser sincero qui su feddit.it ho visto poche vere e proprie spinte aggregative o tentativi d’indirizzamento o spronamento a livello di gestione.
Eppure inizialmente sembravano esserci diversi spunti di riflessione ed idee: per lo più accantonati mesi fa con “ci stiamo pensando” o “è in progetto”.

Altre note sparse che posso fornire da utente ignorante, riguardo la bassa attività:

  • Quasi tutti gli articoli che vengono postati qui - come anche quello in OP - tendono ad esser titoli con due parole in pesca d’interazione ed un link.
    Capiamoci, non c’è nulla di male, però sappiamo tutti che la maggior parte delle persone non apre gli articoli e questa micro barriera è abbastanza per limitare la già scarsa conversazione. Anche solo un bot che fornisce riassunti automatici aiuterebbe non di poco - come si può vedere in alcune comunità esterne.

  • Gli italiani nel fediverso non mancano, ma per lo più son bilingue e, quindi, tendono ad iscriversi ad istanze anglofone più popolose. Questo limita ancor di più la visibilità dei contenuti locali di questa istanza che - con pochi voti ed esigui commenti - non giungono praticamente mai nei feed generali esterni.
    Al contrario ad esempio di quelli Tedeschi o, per un periodo, Brasiliani.
    Per gli iscritti qui basta saltare sul feed locale, ma per gli esterni non è una possibilità facilmente fruibile. Bisognerebbe quindi - se davvero interessa - trovare modi per rendere più semplice scovare i contenuti italiani (come indici, megathread, annunci et similia), ma ciò richiederebbe un grosso e costante investimento di tempo ed impegno nella gestione: difficilmente giustificabile, data la natura senza scopo di lucro e le inevitabili lunghe tempistiche per ottenere i primi risultati apprezzabili.

Ora, come probabilmente è chiaro, la mia esperienza è principalmente limitata al Threadiverso: non ho, e non sono interessato in, account Friendica o Mastodon, quindi non posso commentare quel lato della medaglia.

Mechanize,

Any foundation model is trained on a subset of common crawl.

All the data in there is, arguably, copyrighted by one individual or another. There is no equivalent open - or closed - source dataset to it.

Each single post, page, blog, site, has a copyright holder. In the last year big companies have started to change their TOS to make that they are able to use, relicense and generally sell your data hosted in their services as their own for the intent of AI training, so potentially some small parts of common crawl will be licensable in bulk - or directly obtained from the source.

This does still leave out the majority of the data directly or indirectly used today, even if you were willing to pay, because it is unfeasable to search and contract every single rights holder.

On the other side of it there have been work to use less but more heavily curated data, which could potentially generate good small, domain specific, models. But still they will not be like the ones we currently have, and the open source community will not be able to have access to the same amount and quality of data.

It’s an interesting problem that I’m personally really interested to see where it leads.

Mechanize,

Give technitium a go, my woes diminished drastically with that.

Mechanize,

I used KDE with X11, an AMD card, and two different sized monitors for the longest time without any issues. It’s pretty plug and play.

With Wayland I did try months ago and it just worked™, but I don’t have a long experience with it.

Unfortunately, I’ve no direct experience with the touch integration.

Mechanize,

The Dark Crystal (1982)

Mechanize,

You should add some mushrooms in it.

Is lemmy.world going to ban trans speech?

Lemmy.world just purged access to a lot of piracy communities across the fediverse for its users. They say it’s against rule 1 of their code of conduct, which says no illegal content. And it’s also getting more and more illegal to post gay and trans speech in the US. Are they going to start enforcing the Don’t Say Gay bill...

Mechanize,

I checked in .world’s mod log and I couldn’t find anything piracy related, except the restoration of the community 29 days ago.

It would be nice to have more information from OP.

Mechanize, (edited )

Questo sistema consente di evitare che una istanza appena nata carichi indiscriminatamente tutti i contenuti presenti e passati

Si tratta di una modifica recente? Inizialmente quando s’inseriva una nuova comunità venivan caricate anche le discussioni più vecchie. Ho provato a cercare nei changelog ma non ho trovato il riferimento alla modifica.

EDIT: Da qui

Fetching communities

If you search for a community first time, 20 posts are fetched initially. Only if a least one user on your instance subscribes to the remote community, will the community send updates to your instance. Updates include:

New posts, comments Votes Post, comment edits and deletions Mod actions

You can copy the URL of the community from the address bar in your browser and insert it in your search field. Wait a few seconds, the post will appear below. At the moment there is no loading indicator for the search, so wait a few seconds if it shows “no results”.

Avrebbe dovuto almeno caricare 20 post?

Mechanize,

Scusa, ho fatto un edit mentre rispondevi, lo copio:

Da qui

Fetching communities

If you search for a community first time, 20 posts are fetched initially. Only if a least one user on your instance subscribes to the remote community, will the community send updates to your instance. Updates include:

New posts, comments Votes Post, comment edits and deletions Mod actions

You can copy the URL of the community from the address bar in your browser and insert it in your search field. Wait a few seconds, the post will appear below. At the moment there is no loading indicator for the search, so wait a few seconds if it shows “no results”.

Avrebbe dovuto almeno caricare 20 post?

Mechanize,

MaxTac? 🤔

Mechanize,

I’m unsure about the keyboard but the caps are from the Drop + The Lord of the Rings MT3 Elvish Keycap Set

Update 2.0s been amazing so far but might still need a few patches... (lemmy.world)

Just saw this exploded cop car driving around after sending not-glados back to delamain. Also had a whole cool car chase followed by a bike crash when taking that one cab slowly back to del though and another took me past the the car I’m driving so huge improvements overall.

Mechanize,

If you are on linux you can find the solution here: Steam forum

Probably you can hex edit it on windows too, but I don’t have it so I’m unsure how or which program to use. Maybe in the discussions there are some tips.

Hex Editecho ‘2C45C6: EB’ | xxd -r - Cyberpunk2077.exe

Mechanize,

AFAIK Mistral does already work in llama.cpp, or am I misunderstanding something? I’ve yet to try it.

Mechanize,

I don’t know why but they are on a separate branch: 0.18.5_release_notes

Reading the commits it seems they want to keep them on join-lemmy, instead of having a changelog/releases file in the github.

They first added the 0.18.5 changes and then deleted the full content of the releases file and put to go to join-lemmy.org/news

AMD Radeon 7900 XTX Achieves 890% Speedup In Generative AI With Stable Diffusion Optimization (wccftech.com)

...Both Intel and AMD optimized their software stacks to get massive speedups in generative AI which has seen AMD's RTX 7900 XTX get higher performance per dollar than an NVIDIA RTX 4080 in generative AI (specifically Stable Diffusion with A111/Xformers). Considering Stable Diffusion accounts for the vast majority of non-SaaS,...

Mechanize,

And then here I am, still unable to run StableDiffusion on my 6650xt because pytorch apparently hates it.

TBF llama.cpp runs like a wonder on ROCm.

All aside the improvements in the AI ecosystem by AMD are appreciated and overall tangible. Let’s hope they’ll keep it up.

Mechanize,

It seems it was done to marginally improve serde_derive build times? And just on x86_64-unknown-linux-gnu?

It feels a pretty weird course of action, even if I can understand his point of view his official stance of “My way or the highway” seems a bit stronger than needed, especially considering the amount of problems - both moral and pratical - this modification arises.

I don’t know. If he really feel so strongly about it the only real option would be an hard fork, but a project of that magnitudo and so integrated in the ecosystem is really not easy to both manage or substitute.

Overall it kind of leave a sour taste, even if - I repeat - I understand it is his time and his decision to make.

Mechanize,

I hate that I’m linking to Reddit, but I’m just reminded of this.

OT, but remember you can always use an archived link instead of a live one.

Mechanize,

This is indeed a great summary. I never really modded it, so that too will be a new experience. I hope mods will works with the GoG version!

I remember the ones for Skyrim needed some kind of patching? Or it was only for the SKSE related one? Aaah, that too will be a journey on itself. Haha!

Mechanize,

I went back and checked and it seems I kind of remembered it right, when it was first released on GOG there was some kind of version incompatibility with the Steam one: Link
You can see it in the SKSE page too, there’s a version specific for GOG: Link

But reading in the gog forums it seems this is not a problem for Fallout 4: Link

We’ll have to try and see! ✌️

Mechanize,

it might have been an automated action triggered by reports.

It could well be, but knowing how terse and opaque the site-wide moderation is on Reddit I fear we will never really know. The important thing is that it is now back and even if it was a clumsy attempt at censorship by some of the fan base, as you hypothesize, it failed.

Mechanize,

I feel these graphs tell a different story:

https://feddit.it/pictrs/image/4a7b618a-4cf3-497b-8fb4-d8212b4b1995.webp

IMHO The people who really cared have left, the one who are on the verge are mainly waiting and the one actively participating with votes are the core fans.

The source is Reddit from user nicePenguin, here’s an archived version (Link), they put up a site too at fp-stats.buzz , I’m not sure about the overall accuracy, but it doesn’t seem wrong at first glance.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • JUstTest
  • kavyap
  • DreamBathrooms
  • thenastyranch
  • magazineikmin
  • tacticalgear
  • khanakhh
  • Youngstown
  • mdbf
  • slotface
  • rosin
  • everett
  • ngwrru68w68
  • Durango
  • megavids
  • InstantRegret
  • cubers
  • GTA5RPClips
  • cisconetworking
  • ethstaker
  • osvaldo12
  • modclub
  • normalnudes
  • provamag3
  • tester
  • anitta
  • Leos
  • lostlight
  • All magazines
    •