Notices by fraggle (fraggle@social.coop)

@cwebber even the most permissive licenses like the BSD or MIT licenses require copyright notices to be kept attached, something I'm pretty certain LLMs do not do

@cwebber we regret to inform you the encryption keys were stored in the backup

@remmy @0xabad1dea Piper is Google's internal clone/rewrite of Perforce using its own infrastructure. They have a single monorepo called google3 that stores almost all of the source code for the entire company's products. I don't know how many engineers that is nowadays, but certainly O(10,000s) all using one repo. To make any change you create a "changelist", kind of like a PR in Git terms. Changelists are assigned as a monotonically increasing quantity. So when the company would get close to some significant epoch like "changelist #100000000" a bunch of engineers would all run shell scripts creating hundreds of new changelists to get the nerd cred of being the one to create that one special changelist number.

@cmconseils

@RikerGoogling ok

@mattly I learned recently that Korean has a specific word for 10,000 of a thing and wondered why we don't have one. It turns out we do - a myriad! I knew the word but didn't know that was its specific meaning

@cwebber

@cwebber not committing fully to Maemo was the biggest management blunder that ultimately killed Nokia

Amused that laptop shipping box not only includes a warning that there's a battery inside but also a warning that there's a coin cell battery inside, in case anyone gets tempted to disassemble the laptop and eat the BIOS battery

@aud @dthompson @cwebber I somehow doubt Christine knows what activitypub is

@cwebber @dthompson sorry my attention span is far too short to read all that, I'm going to need someone to generate some kind of slop diagram / political cartoon that can explain it for me

@nikitonsky probably pretty important so that they can be used eg. as part of keys for a dictionary lookup

@cwebber that's why I have this in my bashrc

minfo() {
info --subnodes -o - "$@" | less
}

@losttourist @cwebber EHLO!

@cwebber You hear maniacal laughter in the distance.

@cwebber I'd add, also definitely don't do it the same week as Brooke Vibber Day?
https://en.wikipedia.org/wiki/Wikipedia:Brooke_Vibber_Day

@cwebber you'll soon grow to love the weird taste of internet ham on your pizza

@evan @Taweret I took that screenshot just before my post. Maybe "AI mode" gives better results?

@Taweret here's another good one - fruits that end in "um"

The best part is that the citation it's basing the answer on is a screenshot of itself. As in someone posted a screenshot of it giving this ridiculous hallucinated answer and now it uses that screenshot as a citation to keep giving it. Grey goo process is well underway

@cwebber my personal fear is that we're going to end up seeing a US Supreme Court ruling that the output of LLM models belongs to the company that trained the model

That's potentially disastrous for open source because suddenly companies like Anthropic and OpenAI would get to claim copyright over/demand money from any projects that have incorporated code they helped generate

We're not anywhere near that point yet but considering how much they spend on training the models, it's going to be strongly in their interests to lobby for it and pour money into funding lawsuits that would get them that precedent