can i talk to an openclaw bot using internet relay chat?
Uncategorized
41
Posts
18
Posters
0
Views
-
like i need you to understand, i haven't even gotten through *setup* because the model apparently does not know how to use tools correctly. admittedly it has less than 1 billion parameters, and i don't know what the hell i am doing, but still.
can we get to the part where it is AI winter again already? this is not even fun. i want to throw my computer and its' very expensive RTX 6000 Blackwell GPU out my window.
-
can we get to the part where it is AI winter again already? this is not even fun. i want to throw my computer and its' very expensive RTX 6000 Blackwell GPU out my window.
i just wanted to put an openclaw on irc as a fucking shitpost man
-
i just wanted to put an openclaw on irc as a fucking shitpost man
and you tell me people legitimately are using this software.
how?
is it really magically better when you hook up claude?
-
and you tell me people legitimately are using this software.
how?
is it really magically better when you hook up claude?
(don't worry, i am running this in a MicroVM under kubernetes, I wouldn't dare give it access to anything I care about.)
-
so i installed it into the openclaw meme thing. and it's not like, doing the stuff it claims it is doing.
like it is hallucinating things like "i updated SOUL.md with xyz"
i seriously do not think this stuff is real now
@ariadne what model did you finetune on? For a 1B model you need something really specialized on tool calling.
-
@ariadne what model did you finetune on? For a 1B model you need something really specialized on tool calling.
@jfkimmes i built an LLM from scratch with transformers kinda loosely following the scripts the qwen people released
the LLM is basically trained on ~30ish GB of mostly furry smut and public Linux IRC logs.
*nods sagely*
-
@jfkimmes i built an LLM from scratch with transformers kinda loosely following the scripts the qwen people released
the LLM is basically trained on ~30ish GB of mostly furry smut and public Linux IRC logs.
*nods sagely*
-
@jfkimmes i built an LLM from scratch with transformers kinda loosely following the scripts the qwen people released
the LLM is basically trained on ~30ish GB of mostly furry smut and public Linux IRC logs.
*nods sagely*
@jfkimmes i am, however, using the 35b parameter qwen3.5 reasoning model for the "thinking" portion of this exercise
-
-
(don't worry, i am running this in a MicroVM under kubernetes, I wouldn't dare give it access to anything I care about.)
i wonder if the problem is that the model i trained is too shit to do anything other than really bad ERP
-
@ska @ariadne Ive thouht of something related, not chatbots but imagine a GPS driving assistant voice in your car giving you directions and feedback, but in the most toxic way possible. An angry swearing voice saying things like: "Your exit comes in half a mile, try not to miss that one, you fucking moron." I've thought that ought to be a funny option to toggle on once in a while.
-
and you tell me people legitimately are using this software.
how?
is it really magically better when you hook up claude?
The key is to realise that the average is so low – we can't all be experts at everything, so we are bad at most things – that a model performing slightly above average at one of the tasks we aren't good at means a majority of users will perceive its outcomes as positively better than what they could do themselves.
To any expert, the model falls very short, as it performs well below its own ability.
-
@jfkimmes i am, however, using the 35b parameter qwen3.5 reasoning model for the "thinking" portion of this exercise
@ariadne Oh, is that a OpenClaw specific feature where you can specify that reasoning traces are generated by a separate model than the actual response? I'm not really familiar with OpenClaw's internals.
-
@jfkimmes i am, however, using the 35b parameter qwen3.5 reasoning model for the "thinking" portion of this exercise
@ariadne In any case: as long as the final response is generated by your trained model it will never make a valid tool call since there are probably about zero training examples of the necessary JSON structure required by the tool handling in your furry smut (this is an estimate that could be quite the way off knowing the furry community but still)
-
@ariadne Oh, is that a OpenClaw specific feature where you can specify that reasoning traces are generated by a separate model than the actual response? I'm not really familiar with OpenClaw's internals.
@jfkimmes yes, you can have it use a different model for planning.
-
@ariadne In any case: as long as the final response is generated by your trained model it will never make a valid tool call since there are probably about zero training examples of the necessary JSON structure required by the tool handling in your furry smut (this is an estimate that could be quite the way off knowing the furry community but still)
@jfkimmes this does explain something: it seems to be able to invoke tools when it is planning, but then those tools do not get invoked in the final step.
so it uses tools to read files when planning, then fails to use tools when executing.
what a fascinating conundrum.
Gli ultimi otto messaggi ricevuti dalla Federazione
-
@ilcomizietto Ragioniere, cazzi lei!
-
Concerti di questa settimana.
11.3 BUCCINASCO: Solid Climbing Club (dalle 21:30)
12.3 TORINO: Sin Control Vintage (dalle 18)
13.3 PISA: Spazio Antagonista Newroz
14.3 LIVORNO: Ex Caserma Occupata
15.3 OVADA: Due sotto l'imbrelloCi vediamo nel mondo fisico?
-
titolo: la naftalina di Eta Beta! 🤪
scattata com'è venuta...
🔁 condivisione gradita
11 Marzo, 2026 14:56
-
@ddelemeny@mastodon.xyz @silverpill@mitra.social The confidence comes from an asymmetry I suspect many non-native speakers will recognize: I can read English much better than I can write it.
When I write in English on my own, I often know, as I'm writing, that something is off—that the sentence doesn't carry the weight I intended, or that the nuance I wanted is somewhere between the words I've chosen. I just don't always know how to fix it. When I write in Korean first and then work with an LLM, I can read the result and check it against what I meant. Sometimes I'll see a phrase and think: yes, exactly that, I didn't know how to get there myself. That moment of recognition is the verification step.
So I'm not trusting the machine blindly. I'm using my reading ability—which is reasonably good—to audit an output that my writing ability couldn't have produced alone. It's an imperfect process, but it's not as unmoored as handing a text to a system and walking away.
Your point about polyglot authors is well taken. The tool works better when the person using it can actually evaluate what it produces. I'd agree that's a meaningful distinction.
-
@cgbencini @andrea_ferrero
da laureato in agraria ho sempre tenuto i piedi per terra e interessato poco a ciò che sopra la mia testa, ma trovandomi a insegnare rudimenti di astronomia e geografia astronomica cerco sempre di dare un'interpretazione logica che vada un po' oltre la mera nozione
-
@hongminhee@hollo.social I think your post makes me think more critically about the use of LLMs for translation services. It is easy for me to judge from a position of privilege because I am a native English speaker, and I do not realize the access I am given simply because of it.
While reading your post it also made me think about the sacrifices you made to contribute to this community. There are precious few people in the Asia-Pacific region who regularly contribute to AP development, and a large part of that is the language barrier.
If LLM-translation makes the AP development community less euro-america-centric, then I am all for it. Cultural differences we can work through, but language barriers are harder to bypass!
Aside, congratulations for making it onto Hacker News front page :)
-
In guerra mandiamoci i vecchi
https://www.ilpost.it/2026/03/11/bellu-in-guerra-mandiamoci-i-vecchi/
> Una modesta proposta per il ministro Crosetto che ha chiesto di “prepararsi per il peggiore scenario possibile”. Gli anziani sono sempre di più e sempre più in forma...
-
@hongminhee @silverpill I really think there could be a lot to do in terms of bridging the fluency gap in terms of UX. You refer to your experience flipping pages of dictionaries, and I relate to that quite hard : that's where I'd like to see effort and change in software.
However, I feel comfortable bearing the responsibility of making my speech accessible to an English or Spanish speaker that doesn't speak French, and any failure would be mine.
