Hi community
Recently, we’ve had people attempt to submit ChatGPT-generated sentences to MCV. We are excited by the opportunities that large language models (LLMs) provide for innovation, and we understand that many of our communities face a significant challenge generating open source sentences.
However, presently, we are not accepting AI generated content. LLMs have, for the most part, been trained on data scraped without consent from the internet, and there are ongoing conversations about whether copyright has been infringed. Even outside of the legal conversation, we are passionate proponents of community-centered data creation, and we worry about a world in which people are left out of decision making processes about their data.
We thank everyone for their patience, we welcome your feedback and perspectives, and we’ll keep the conversation open and ongoing. If you’re looking for resources, ideas and support for creating sentences for your language, check out some of our ideas on gathering, validating and processing public domain sentences.
