What's the latest issue regarding large language model training, according to the NY Times?
- OpenAI has harvested data from YouTube audio transcripts
- There is a growing shortage of good quality data
- Google is harvesting data from Maps and Reviews
- All answers are true