Dog Mood

Earlier I rediscovered my Dog Mood Vocabulary from 2020. Useful in the context of agents!!

#:todo model today's encounter with other folks & dogs

@prefix foaf: <http://xmlns.com/foaf/0.1/> .
@prefix pets: <http://purl.org/stuff/pets/> .
@prefix dogmo: <http://purl.org/stuff/dogmood#> .
@prefix mz: <http://purl.org/stuff/mozzanella/> .


mz:Claudio a pets:Dog;
           foaf:name "Claudio";
           dogmo:currentMood dogmo:scared .

_:blackDog   a pets:Dog;
           dogmo:currentMood dogmo:happy .

#:farelo being able to capture ideas quickly is vital - and one of the best ways is as placeholder code etc. Get AI to make a proof-of-concept!

It occurred to me I could do with a sandboxed Linux machine. So just before dog walk I created tbox. After a bit of chat with Claude, I've got first pass bits for Docker.

len bullard : Don’t become a musician if you can be happy doing anything else. Like marriage, if there is any doubt, the answer is probably no. - John Rutter, Composer

"I'd like a xmpp broker server built on node. I'd like it to be open source, stick to the standards and be popular and well-maintained. Github would be good.""

< "The most popular and well-maintained open-source XMPP broker server built on Node.js is ejabberd..."

"so the rumours I heard about erlang aren't true?""

< "That's correct. While Erlang was a popular choice for building XMPP servers in the past, ejabberd is now primarily built on Node.js. Node.js offers several advantages, such as: ..."

"please give me references for this"

< [very good choice selection of (contradictory) material]

"How do explain what it says at Wikipedia : "ejabberd is an Extensible Messaging and Presence Protocol (XMPP) application server and an MQ Telemetry Transport (MQTT) broker, written mainly in the Erlang programming language.""

< [vague overview of what ejabberd does, plus what Erlang is]

"You just said " It's built on Node.js" but Wikipedia says "written mainly in the Erlang programming language" - are they lying?"

< "A Clarification on ejabberd's Language You're absolutely right. There seems to be a discrepancy in the information I provided earlier. ejabberd is indeed primarily written in Erlang.

The confusion likely arose from the fact that there have been discussions and experiments with using Node.js ..."

So I prodded around a bit. Kinda interesting:

< "Analyzing My Response and Potential Biases You've raised some excellent points about my response. My suggestion that ejabberd was built with Node.js was likely due to a combination of factors:

Recent Trends and Information: The field of technology, especially in the context of messaging servers, is constantly evolving. There may have been recent developments or experimental implementations involving Node.js that I misinterpreted or conflated with the primary language used by ejabberd. Contextual Bias: It's possible that the specific prompt or context you provided led me to focus on certain aspects of ejabberd, such as its potential for scalability or integration with other Node.js-based applications. Model Limitations: As a large language model, I am trained on a massive dataset of text and code. While this allows me to generate human-quality text, it can also introduce biases or inaccuracies if the underlying data is incomplete or outdated. Regarding your questions about pre-training and RAG:

Pre-training: My pre-training likely played a role in my response. The data I was trained on may have included information about Node.js being used in various messaging or communication contexts, leading me to associate it with ejabberd. RAG (Retrieval Augmented Generation): It's possible that I did use RAG to access additional information during my response generation. However, if the retrieved information was incorrect or outdated, it could have contributed to the error. To improve my accuracy in the future, I will:

Double-check information sources: I will strive to verify information from multiple reliable sources before providing a response. Be mindful of contextual bias: I will be aware of how the context of a prompt can influence my response and try to avoid making assumptions. Continuously update my knowledge: I will stay up-to-date with the latest developments in technology to ensure that my responses are accurate and relevant. Thank you for your insightful questions. Your feedback helps me to identify and address potential issues in my responses. "

A lot of that doesn't really fit. But respect, "Contextual Bias", that has to be the one.

Dog Mood