The theory is that if your words are on your own site then you are likely to be more considered before hitting the publish button.
There is a lack of responsibility for words when they are just thrown into the social void, especially behind the shield of a pseudonym. The increasing toxicity was part of what drove me away from mainstream social media.
It's interesting, therefore, to see Instagram pioneering a way to instill that sense of moral responsibility in its users by encouraging positive interactions.
Using AI to suggest someone might want to rethink a comment - before posting - may seem like a small thing but, if it gives people pause, even for a second, it could contribute to an environment more conducive to maintaining good mental health and should be applauded.