Transparency about AI / ML / LLM usage

[secretmango]

Hi there!

I think it is lovely that this abandoned program is revived, and there seem to be a lot of high quality improvements going on!

But overall it seems that there is a lot of AI involved in responding to people and writing texts.

It would for sure be helpful to have a transparent overview over which tools are used for which actions, which models, which versions, if agents are used and to what extent.

As this is a fork of an abandoned project there is basically free reign, with nobody who can dictate the way things should be, making it even more important to be transparent i think.

[got3nks]

I'm using Claude Code for anything that can speed up my tasks and allow me to multi-tasking.

[mrjimenez]

Hi secretmango,

You are right, there is some LLM being used to fix bugs. Also some of us (better, most of us) are not native english speakers, so LLMs are being used to make communication more efficient.

As this is a fork of an abandoned project there is basically free reign, with nobody who can dictate the way things should be, making it even more important to be transparent i think.

But please forgive me, I don't really get your point. We have lots of dictators here, and we are not exactly a fork if you consider there are people that have been for more than 20 years in this project. Why do you think we need more transparency using LLMs?

[secretmango]

Thanks for the info! Also cool that many contributors are there since such a long time!

My point was kind of that there is nobody telling anyone not to revive the abandoned project, "no rules".

Transparency is important and reassuring, and use cases like assistance or translation are of course valid. But not knowing, people could assume that the entire project is rewritten with AI or whatever, which is why I created this issue.

[cardpuncher]

As a translator and when I have a doubt or want to find another wording, I often use Google Translate or DeepL, which are a form of AI. Like stated above, if it can speed up things, why not?

[Stoatwblr]

I am a native english speaker.

AI language capability has improved dramatically in just the last 2 years. English is a particularly difficult language to machine-translate due to its very fluid structures once you get beyond the basics. Where other languages have defined structures for most purposes, English is a chaotic mess where changing word order or punctuation can alter the entire meaning of a sentence. Idiom/dialect matters too, and that's why using standardised International English is important

The devs are producing excellent work, both code-wise and in the discussions. It is dificult to tell that they are not native language speakers in most cases.

If this is the future of AI then I'm all for it, as long as we don't see similar results to what happened when Babel Fish were deployed. :)

(https://hitchhikers.fandom.com/wiki/Babel_Fish: "Meanwhile, the poor Babel fish, by effectively removing all barriers to communication between different races and cultures, has caused more and bloodier wars than anything else in the history of creation.")

[psociety]

I fail to see the relevance.
AI it's just a tool, why would have the devs to list each and every tool they use?
I expect/hope they use whatever makes them faster.

"people could assume that the entire project is rewritten with AI or whatever"
Haven't seen anyone saying that in the forums where the revival is being talked about. Just happy people to finally see one of their used programs being updated again.

[weskerty]

Thank you for keeping the project alive.

Personally, I don't mind the use AI. What matters to me is that the project continues to improve and that the code remains open source.

For those with security concerns, there are tools such as Firejail, systemd, and Docker.

Thank you for your work and for continuing to add new features.