Egregoros

Stayed up until 1 am developing a 20 year old hentai game nobody cares about again award

@solarpapst@social.anoxinon.de Don't you just love it when your country's military launches the equivalent of all taxes you will pay in your entire lifetime into a single missile to kill some random people in a desert that were never a threat to you?

@mischievoustomato@tsundere.love @phnt@fluffytail.org @sun@shitposter.world You're never "wasting time" when you're hacking, that is a false narrative created by the proprietary business oriented overlords.

If you didn't achieve something, you still learned something. You didn't "waste time".

@mischievoustomato@tsundere.love @sun@shitposter.world Doesn't it make more sense to just not use it until then? Can your rotten brain not live through a little bit of FOMO?

@mischievoustomato@tsundere.love @sun@shitposter.world You're not kneecapping yourself by not using something. These services didn't even exist a couple of years ago. Did you feel kneecapped back then?

@sun@shitposter.world I indeed want to have those things too, but I will not tolerate the SaaSS ones in the meantime.

I personally don't think LLMs or GenAI software in general is always unethical by definition.

Currently almost all available implementations of it are unethical, but I see this as only temporary. There are definitely ways to create these AI models in an ethical way. We just haven't done that yet.

We also haven't fully figured out all the requirements to even make them respect the user's freedom in the first place.

I'm getting kind of frustrated because mainstream opinions are either "GenAI bad, you must never use it ever at all" or "GenAI best thing ever ever happened to computing, we will reach AGI soon".

Not many people looking at the current tech hype in a more objective way.

@jeffcliff@shitposter.world @sun@shitposter.world @yonle@fedinet.waltuh.cyou The local LLMs I have tried are pretty decent at writing test classes. I think that's probably one of the better usecases.

Tests are a lot of work, lot of boilerplate bullshit, and you don't let the LLM actually change the actual code. And it's very obvious when it mistake.

@yonle@fedinet.waltuh.cyou Vibecoders are ridiculously dumb and lazy. Do you think they actually know how or care enough to remove claude's coauthorship on their commit?

@yonle@fedinet.waltuh.cyou People will literally be wasting their tokens doing stupid shit like:

"Hey Claude, can you commit these changes?"

Instead of just typing

git commit -m 'changes'

@sun@shitposter.world Are Amiga 500s that expensive? I fixed one up at the hackerspace.

@sun@shitposter.world I have a Debian Squeeze virtual machine to play with Python 2 stuff.

@xarvos@outerheaven.club @lizzy@social.vlhl.dev
LLM usage in official GNU projects?

I think I'm gonna puke.

Is nothing sacred anymore?

@xarvos@outerheaven.club @lizzy@social.vlhl.dev I have a huge urge to reply to this email thread but I am neither a Hurd user or developer so really not sure if intervening with my personal opinions about LLM usage is what I should be doing with my time.

@Pi_rat@shitposter.world I think "mansplaining" is a very sexist term.

You can tell someone you think they are being too condescending in tone without making this about their gender.

It happens quite often that some misandrist woman on here complains that I am "mansplaining" to me. So I'm going to explain it again here.

I am not "mansplaining". I talk like this to everyone. This has nothing to do with your gender. I am just autistic. Know the difference.

@sun@shitposter.world @lizzy@social.vlhl.dev TempleOS is perfect and if it doesn't work it is clearly your hardware that is wrong.

​​

@lizzy@social.vlhl.dev Really not a problem.

There's already a bunch of alternative kernel projects like GNU Hurd, FreeBSD, Illumos, Zircon, or even Plan 9, etc. that are all waiting on the sideline like;

"I can't wait for Linux to collapse so MY kernel can rise from the ashes"