I would bet that 88% accuracy is relatively on-par or better than humans perform at the job. Pretty fucking cool use of “AI”.
https://uwaterloo.ca/news/media/ai-saving-humans-emotional-toll-monitoring-hate-speech
I would bet that 88% accuracy is relatively on-par or better than humans perform at the job. Pretty fucking cool use of “AI”.
https://uwaterloo.ca/news/media/ai-saving-humans-emotional-toll-monitoring-hate-speech