Alternatively, how many of them have invested in one or more of these LLM makers and are ready to torpedo their own business as long as it makes the share price go up/feeds more authentic training data?

not far off, sadly

tinfoil hat on: discord is a company based in the USA, the USA administration saw those teenagers in Nepal supposedly organise a revolt/revolution over discord, sees the inhabitants of Minneapolis organizing against ICE on Signal, and wants to preemptively know which of their citizens are using discord with which accounts in case any of them start organizing on discord.

For those curious about how this fork came to be, KnowYourMeme had a decent rundown: https://knowyourmeme.com/memes/events/godot-engine-user-blocking-controversy-wokot

Pretty easy to sum up in 1-2 sentences…

Then by all means, give them your 1-2 sentences per DE so that they “only” need to include them!

Frankly, I think it’s a lot harder than you’re making it out to be, especially over such a large range of DEs. Not that the suggestion is without merit, just that the assumed difficulty of making it work as intended (i.e. actually helping a new Linux user pick the “right” desktop environment for them) seems underestimated.

Maybe Cinnamon can get away with “it’s like windows 95”, but Gnome and i3 are quite different from anything the target audience has ever experienced.

Then I guess it’s time to put “AI” (actually 3 if-statements in a trench coat) into all my software projects so they can legally jailbreak corporate software!

We might finally get a triumvirate of generalist, centrist instances!

…or .world will crash, burn, and implode

Speaking of which, nice username

…

…

though I’m not sure I get the “train” bit

At least they aren’t green tips!

Good for them! I’m not surprised yet still disappointed by the spokesperson’s comments.

As well as 200 miles from every international airport inside the US.

Given the stochastic nature of LLMs and the pseudo-darwinian nature of their training process, I sometimes wonder if geneticists wouldn’t be more suited to interpreting LLM output than programmers.

I see, thanks for the correction.

There used to be this website, but the url just loads up a scam site now (I’ve created this issue on the project’s tracker if anyone has additional info to contribute).

I don’t know how technical you are, @VieuxQueb@lemmy.ca , but you could try running the “defed-investigator” project locally.

lemmy.ml, no, but I’m fairly certain that lemmygrad.ml has been defederated from lemmy.world at least, if not others.

I just searched on GitHub for "Sha1-Hulud: The Second Coming.": 692 repositories. On the first page of results I was able to find a repo clearly made by the malware, and in that repo I was able to find someone’s github token with a few applications of “decode from base64”.

This is pretty bad. I don’t know what exactly comes next, an awareness campaign to get people to clean their infected machines and packages?

According to this article written in July, it’s a bit more dire than that if you take a step or two back. Basically, openai and their copycats/derivatives are being held up by investments from Microsoft, Google, Amazon, and Meta, who in turn are being held up by investments from Nvidia. If/when the whole chain collapses it’ll be more than 0.5% of earnings that disappear.

Not that I disagree, just as someone who loves computers and programming it really feels like throwing out the baby along with the bathwater.

We could (should imo) be planning a sort of overthrow of the rich assholes who don’t share; make sure everyone has access to a computer, the electricity need to run it, and the knowledge to use it to their own benefit.

The second, longer quote in my previous comment is from the intro to a computer self-help/“how-to” book, Without Me You Are Nothing (pdf link).

“Once men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to enslave them.” – Frank Herbert

"Right now there is an explosive growth of the number of computers and things they can do. Not only are their numbers increasing at a dazzling rate, but the storage of information in giant data banks is growing in the same explosive way.

We have no way to control this now and none in sight. In fact, the very nature of this growth says that all controls will lag far behind computer developments. Any attempt to ban them will only drive com- puters underground. Never lose sight of the fact that computers “crunch time.” The speed at which computers can operate tells us that laws cannot keep up with them. The person with a computer can dance rings around you while you react as though you were embedded in molasses.

What can you do?

Get your own computer. Learn how to use it. We are here to help you make that first step: how to find the one that fits your needs and your pocketbook, where to put it, how to program it-all of the essentials. If you don’t do this, the Bill of Rights is dead and your individual liberties will go the way of the dodo." – also Frank Herbert

I hate how much we seem to be slowly careening towards Frank Herbert’s vision like the worse case of collective target fixation.

Incredible. All throughout my studies the “bitter lesson”, so to speak, was that analogue circuits just couldn’t hold a candle to digital ones in terms of reliability when operating on small currents, to the point that no one bothered to miniaturize in analogue anymore. Even guitar pedals are almost all transistor-based, because it’s so much more feasible to to manipulate small currents through binary, quantized signals than analogue ones (even though the analogue ones are theoretically infinitely more precise).

Here’s the publication in Nature: https://www.nature.com/articles/s41928-025-01477-0

Here is either the pre-print or an accompanying paper on arXiv: https://arxiv.org/pdf/2205.05853

I’m trying to figure out how they get around the compounding imprecision that is inherent to multiple analogue steps and actually manage to rival digital circuit’s precision; seems like a big part come from how they have managed to squash all of the useful “work” down into almost a single step thanks to clever use of operational amplifiers on the “edges” of their resistive random access memory array.