Artificial Intelligence Can Now Spot Arguments Before They Happen

Machine learning, a form of artificial intelligence, is developing the ability to spot arguments online before they happen, and gives you a gentle warning that things are heating up.

This could be a good thing. We all know social media has devolved into a divisive, argumentative, cesspool of people flinging mud balls at each other from the safety of their keyboards. Anything that can help tone down the nasty exchanges I see everyday is welcome by me.

Except I think the machine learning might be a little too sensitive.

For example, today on Facebook I wrote, “I love the fresh-squeezed orange juice at Sunny Sally’s restaurant over on Pico.”

I received this message within minutes. “Hi, it looks like you might be threatening to throw juice on Sally. Please take a moment to consider your words.”

What?

And later I wrote, “The Warriors will crush the Rockets into submission with their sledgehammer of talent.”

This time the message was immediate. “Hello, violence will not be tolerated on Facebook. Please choose better words in the future or your account could be terminated.”

So I wrote, “A bot just threatened to close my account. LOL.”

There was a knock at the door. Opening it, I saw a robot standing quietly. A whirring sound emitted from deep inside and it began showing a hologram of Mark Zuckerberg on my porch. “Seriously, dude, cut with the sledgehammer talk,” he said, glowing slightly.

“Sure, Mr. Zuckerberg,” I said, not knowing what I would do with my life unless I checked Facebook 189 times a day.

Zuckerberg looked away and then back at me. “And the Rockets will win tomorrow and take the whole thing from the Celtics,” he said.