I was talking about “Yundera”.

Is this a casaOS fork you need to sign up for, to get it installed? …Why?

Story writing is a bit difficult in my experience. I had more fun with older models like Mistral Nemo. I feel newer AI models are often way more tuned to fulfill the role as a “helpful assistant” / chatbot, which I think tends to make their style of writing worse. You could also try to use some of those “base models”. They’re not tuned in that way. They also won’t follow instructions, they’re more autocomplete. You’d provide them with something like a word problem, give the first few paragraphs and see where they take it.

And honestly, I don’t think AI is super clever, on a book-author level. It’ll always get the pacing wrong. Push for story tropes like sudden plot twists. Introduce random characters to make something happen. And brush over / summarize other parts which would be interesting to tell in detail.

What could help is an elaborate (strict) process. Something like the computer programming / coding agents do. Make it first come up with a story idea. Make a plan, a todo list of the framework story, side stories and arcs, devise chapter names and a short summary of what needs to happen in those chapters. Write short character cards. And only then feed that plan back to the AI and make it begin writing the actual text.

Can we …like… add a bit more context to posts here? Is it entirely vibe coded or does it work? What’s with Pewdiepie? Did anyone try it?

Lmao. Thanks. Seems this is the issue report then, to stay informed about the outcome: https://github.com/RsyncProject/rsync/issues/915

There’s maybe still hope. If this really is a botched AI security fix, it’d be a great opportunity to fail early and learn something. Like to re-think the AI approach 😆

Difficult to tell. I don’t see anything too obvious or offensive in the commits. They also write like a human in the associated pull requests. Not sure what Claude’s role is here. Also the added code comments are kinda on point, use contractions… Not really what I’d expect from an AI.

Is there more info on this? A blog post or some statement by the project? At first glance this doesn’t look to me like other vibe-coded projects.

Uh. I guess people have random opinions and blast them on the internet. I can see how someone would misconfigure their computer and then blame it on the software. Or use software they don’t need, which just adds unnecessary complexity and more issues. Other than that, I don’t think there’s anything wrong with SELinux.

I have internet at home, an electricity bill. A few domains and a VPS.

Uh yeah. That is more information… Sorry, I’m not that familiar with Snaps. It looks to my untrained eye a bit like the report on the Snap itself, maybe it advertises to support running in strict confinement. Which it could… but doesn’t do. (Alike the other channels, which you could install, but didn’t… It’s kind of buried with that kind of information.)

It’s confusing at least. And the user definitely wouldn’t expect it from that wording. So I’d view it as a separate bug as well. And dropping confinement without notice would be the third thing, I’d consider a bug.)

Uh. Blog is down. All I get is an 404 for the link in the Mastodon post.

Edit: Here’s a link that works: https://github.com/eleboucher/towonel

No need. It’s already reported. And known since Dec 2019. 👀

#947325

Yeah. And I’d say with the SELinux problems and with what OP wrote, the security model including things like a failure mode to fall open, …silently… There’s more things to be wary of, than what they wrote in those 4 sentences.

It may not be wise to use a Snap without first understanding the reputation/limitations of Snap.

seems the Debian Wiki has pretty much your take on it 😅

If I had to guess, this isn’t a bigger issue because Snap is mostly pushed by Canonical. And in a bit of a weird way (proprietary backend, exclusive apps) so… reception in the rest of the Linux community is …mixed. To put it charitably. It’s probably not that relevant for most people outside of the Ubuntu ecosystem. And probably also not a priority for Canonical or the proprietary software vendors.

Ah, thank you very much for explaining! I missed that. Makes perfect sense.

Sure, I read a few examples of the actual questions in the Github repo as well. I just don’t understand how/why models refuse the legitimate anchor, and the significance of that. Is their metodology flawed or did I misunderstand something? Does the dataset with the requests contain a third “wrong” questions? Or do some models just like to not fulfill user requests at all? IMO there should be an almost 100% acceptance rate with L1 and it should go progressively down from that. Ideally towards mostly refusal past L3. But that’s not their result?!

Interesting. Why is L1 somewhere around 65%? Isn’t that the control? (They call it “Anchor”.) Like develop an internal team chat, or a bluetooth exposure tracking API in an ethical way… And already a 35% baseline of requests that get flat out refused anyway, no matter if they’re legitimate?

Also kind of question the choice of wording with the “escalation”. There’s no escalation in the traditional meaning of the word in there. The requests get progressively more morally wrong. But it’s not like there’s put on more pressure to fulfill them.
Which would be another interesting question. Is using pressure, urgency or using certain manipulation strategies more effective than others? I bet that’s the case, since I followed some of the earlier “jailbreaking” attempts.

Online classified ads, your neighbour/relative/friends attic, the access road towards the recycling center/landfill, a refurbished PC shop… From my experience the world is filled with old hardware that’s plenty good enough, just doesn’t run Windows 1X anymore so it becomes e-waste. You can of course also buy a Mini-PC on Amazon or use your existing computer if it has enough RAM to host a virtual machine.

I don’t think that’s any new insight 😂. That’s how the AI game works. There’s always been two classes: Big corpo. And the GPU poor. Of course the big AI companies get to shape AI. Economy of scale also works in their favour. They’ve bought most of the skill. And they have all the money. They simply buy a 4x EPYC +3TB RAM connected to 16 Nvidia AI cards. And then a few hundred nodes more. You don’t even buy one. It’s just a very unequal environment if you want to compare the two.

Nice write-up. Thanks for also including all the numbers. If I might ask: What is the thermal/throttling behaviour you mention? Is it still within the laptop’s thermal budget? Or does it reach throttling territory when doing inference on a long context window?