Right one!

Well just create another service with mounted paths and run shell script inside it or use systemd run --property ...?

What do you meam by outside of service, but within systemd?

That was the whole point of locking down everything and making it illegal to fix by yourself.

It will bite them in the ass when next generation would have a lot less people who can work for these companies (it’s already happening).

python 2

🤡

4x8tb they had 8.5k hours on them when I got them four years ago, they work non-stop since.

There is nothing wrong with modifying the model to make it look better, printers print differently, end results look different. It would be a benchmark still, as in you obviously can print better, you just need to make necessary modificafions.

Benchy license apparently prohibits derivative works. hackaday

Thanks, to be honest angband was pretty easy as long as you have an understanding of how to play rl. Brogue has some similarity in this with nethack, you need to know and remember things from previous games, which is hard for me. 😅

Infra Arcana was the brutal one due to a lot of clever machines on levels and sanity management and it’s pretty unforgiving in general.

I don’t think I’ve beaten any other completely.

brogue, not very good at it and not that interested in reading spoilers to learn.

Thank you! Ai tagging sound like a super cool feature hope it works well enough.

Tool calling is cool funcrionality, agreed. How does it relate to openai blowing its own sails?

It works as expected, they give it system prompt that conflicts with subsequent prompts. Everything else looks like typical llm behaviour, as in gaslightning and doubling down. At least that’s what Iu see in tweets.

There is artificial, there is no intelligence, there is a whole lot of artificial neural networks without any intelligence. As in calling them ai in the sense that these are thinking machines comparable to some animal (human for example) is misleading.

Do you disagree?

They written that it doubles-down when accused of being in the wrong in 90% of cases. Sounds closer to bug than success.

There is no ai in ai, you chain them more or less the same as you chain browser or pdf viewer installed on your device.

"Open"ai tells fairy tales about their “ai” being so smart it’s dangerous since inception. Nothing to see here.

In this case it looks like click-bate from news site.

What? Jellyfin doesn’t need any “suite”? You can use it with your collection obtained whatever way.

You can buy vps in china that would be locally accessible, for homelabs and the like there are tools like frp, rathole. (I am not from china)