According to "The algorithms that detect hate speech online are biased against black people" by Shirin Ghaffary, what is a reason that AI models incorrectly flag tweets as

hate speech?

Current computing resources CANNOT implement the complex Al algorithms to process tweets

Al works slower than humans in reading the text of tweets

Al has a difficult time understanding social context which is important for understanding when a word or phrase may be hate speech and when it may not be

Al is NOT scalable to process the quantity of tweets people can produce