This post was split from another regarding the usefulnewss of Sutta Central, Bhikkhu Bodhi, and Ajahn Thanissaro. They have all created useful translations and done great things. The problem is there are some different views expressed that are not CT which are easily seen to confirm their views as CT. After all they are Theravāda monks.
Yes… suttacentral is also very useful, which goes to the “australian brothers”. We even has their translations built into Tipitaka Pali Reader so one can read pāḷi/english. This was actually my own idea that I implemented. The problem is SC organizes a lot of data very well for re-propagation and it is CC0 too. The really cringeworthy part of all of this, is that AI will become the way that dhamma is spread, and since SC gets many points for popularity and maybe “perceived reliability”, it could be more influential than we might believe on the surface. The good news is that AI does indeed know other Theravāda languages including Pāḷi itself. It is not perfect, but it should be able to sort through everything better in time. However, the Sinhala and Myanmar data is not as widely digitized, but the audio dhammatalks exist.
So I guess Ajahn Sujato does not like the idea of AI using his CC0 published work…
That is surprising. His data is king and ready for anyone to grab. I reluctantly use that data and I have considered AI using his data a few times and the negative affects of his version of dharma getting spread to AI which is surely the future. (notice I called it dharma and not dhamma). I’ve always been worried about this and for different reasons, Ajahn is worried about this too.
In all fairness the SC translations are not so bad. In some sections he follows the commentary definitions… For instance “only eating in one part of the day” rather than “eating one meal”. This is a suttanta-commy-basher’s favorite and there he goes and translates this well. Some words like prudent and astute really irk me… but that is the price to pay. It is cc0 so I guess I can change it if I want. And that is the nice thing about cc0
take a brahmin, merchant, or worker who shaves off their hair and beard, dresses in ocher robes, and goes forth from the lay life to homelessness. they refrain from killing living creatures, stealing, and lying. they abstain from eating at night, eat in one part of the day, and are celibate, ethical, and of good character.
If the irk is about changing the well established key concepts to something only slightly more accurate, I agree. I will change my own internal definition of “heedfullness” to something that involves a bit more prudence. But I don’t want to start using a new word where I already have a billion associations to “heedfullness.” This is one of the places where a respect for tradition is valuable.
The good thing is that is cc0 and you can change anything you want.
They are going to change their policy soon and may try to re-copyright their stuff so that ai does not use it. I’m not sure I have the originals anymore. I should probably take a look at the copyright again and then fork the project. Why you cannot undo a cc0 declaration, you can make the data difficult to find. it would be good to make a list of substitutions that can be made and try to figure out if the english will work. I had an idea of actually having a sw program that gradually replacing the pali back into the english as one chose. Therefore using as a tool to learn pali with the words they are familiar with.
I LOVE that idea.
Im conceiving an interface where these can be ephemeral, fading in and out over a defined number of seconds (slider) translations (multiple select) and destination language (checbox).
My biggest irk is “choice” for “Volitional Formation.”
That one is just… whatever. Vatican II comes to mind. But it’s more than simply bland, it toses out the whole “formation” bit. There’s a REASON that’s in there.
Bhante Sujato hasn’t mentioned anything of the sort. The license statement has always included the explanation that he hopes that the work will be used in line with Buddhist principles. All he is doing now is asking that people not use the content to train AI as he believes it is not ethical. He has written in his recent posts that he acknowledges that request is not a binding part of the license.
Sorry if this isn’t quite on topic, but since people here probably don’t spend much time on the SC forum, I thought it would be good to share what I know.
That’s a neat idea. Bhante Sujato’s translations are remarkably consistent so this should be easy to do as far as the text. In fact, I’ve experimented with this using a browser plugin like this one to replace words in the pages. For example switching “choices” back to “sankhara”. It’s been a while since I tried, but if a plugin could be found that allowed exporting and importing of the substitution list, then you could even maintain a file here on CT for people to transform Bhante Sujato’s translations into what you see as more Classical Theravada.
If someone believes in good faith that AI is good and will do good for the world, then it will be within the intentions. But public domain is public domain.
I do a lot of work with Tipitaka Pali Reader which I had plan to make a gradual replacement of the pali to english. There are ways to do this. I just don’t have a lot of time… especially now. About 90% of the changes in the past 6 months have been made by GPT+ actually. It knows how to program very well.