You are viewing a single comment's thread:
Thanks for your feedback @palomap3. It is very much appreciated.
I also see the issue for users that comment on content that is not in their own language. There is hardly any Dutch content on Hive, so I'm always writing in another language. I also use is sometimes to form a good sentence, but I do think there is a big difference between "exactly translate my comment" and "write me a comment in language xxxxx". I seem so see more of the latter lately.
For the point of replying in the language of the blogger; I personally don't have a problem with replies in Spanish (or any other language). I appreciate the time you have taken to engage with my content. This is an invitation to comment on my posts in Spanish ;)
I am using the Chrome translation. I've had some nice Spanish/English conversations on some of my blogs before.
But I can't speak for others...
You mention you are using a AI to translate, but I have never had any doubts with any of your comments. AI comments have certain patterns I recognize quite well after seeing a lot of comments.
Yes, we are going to have some 'false positives' of well willing people. The good suffer because of the bad.
Just an open question; would it help if we did notify the commenter of AI usage?
I have done that in one case and the 'culprit' did say they used it as a translator.
I'm not saying we will do this, but that might give the hivers that are unaware of this an option to reply.
(Se me hace muuuy extraño responderte en español, que conste!) 😁
Entiendo perfectamente vuestra intención, yo tampoco tengo dudas de ninguno de vosotros. Sé lo que es curar contenido, he dedicado cientos de horas a esa labor y he visto de todo. Mi única apreciación es hacia el baneo de los usuarios. En lugar de tener una lista negra de infractores, ¿sería viable una lista blanca de los que sepáis que son auténticos? Es una pregunta al aire, no hace falta que respondas, es una sugerencia.
En cuanto a tu pregunta final, no lo veo positivo para el proyecto, daría más problemas que soluciones. No tenemos una forma verdaderamente objetiva de demostrar que algo está escrito con IA. Quien la usa para abusar lo ignorará o lo negará, me ha pasado con varios casos. Ante la duda mi opción es pasar de largo o enviar informe al proyecto que recibe fondos para lidiar con eso (HiveWatchers). Por otro lado, también hay falsos positivos. Yo misma he pasado textos míos por los detectores y los he recibido. 😳 ¡Y sé bien lo que escribo y cómo!
View more