Don’t get fired over an angry text

Let artificial intelligence help you out

It’s Friday 11 pm. David is struggling with a deadline at work.

His friends have already started the weekend party. Unfortunately, due to last-minute client’s demands, David had to stay back crunching numbers.

Suddenly, the screen blinked. It was his client, Jeremy.

Jeremy, “Hey, don’t mind me saying this but your team has done a really lousy job with the presentation, I suggest you re-work on slides 5,7,9 and 11.”

David, “Sure, I’m on it.”

Jeremy, “And while you are at it, also make sure you proofread the articles. This is urgent! You guys have already made me re-work the slides for the last four iterations, the least you can do is wrap it up today?”

David was infuriated. He knew he had done a good job with the presentation, and the rest were minor cosmetic changes. Jeremy could have easily fixed it himself. It was already a hectic week for David, and all he wanted was to join his friends right now for the party.

Suddenly, David’s fingers blasted on the chat, typing harsh words to Jeremy. He was tempted to tell him that he’s a jerk.

Enter: Tone detector

As soon as he hit the send button, the screen prompt appeared:

Are you sure you want to send this? Do you wish to take a break?

David realized the bot was right. He needed the break. It was unfair to blast on someone who was just another part of the system.

The scenario is not all that alien. Most of us have been in a more or less similar situation one time or the other. Statistically, angry texts are one of the major reasons people get fired.

In a recent report by Gartner, 26% of 4000 surveyed mobile users admitted that they faced severe repercussions at work (including getting fired), because of an angry text. 81% mentioned they had sent a message they wished they could take back.

Thanks to artificial intelligence, you can (now) analyze your word choice, phrasing, punctuation, and even emotions before hitting send.

What is tone detection?

In simple words, tone detection helps you understand how your texts are sounding. Also known as opinion mining or sentiment analysis, it’s goal is to uncover and analyze the emotions behind your texts.

How does it work?

Tone detection uses linguistic analysis to detect human emotions (like joy, fear, sadness, anger) found in the text. There are plenty of ways to analyze sentiment, but the most popular two are the lexicon-based model and machine learning.

How can you benefit?

While tone detection has already helped industries like customer service to improve the quality of interactions, office communication is still an unexplored territory.

Many big players in the workplace collaboration space like Slack, Microsoft Teams, and Flock are still not sure about prioritizing this feature in their product roadmap.

Grammarly is an app that automatically detects potential grammar, spelling, punctuation, word choice, and style mistakes in writing. It’s quite non-intrusive. It won’t stop you from sending angry texts. It would merely prompt you.

Imagine an AI assistant in your office chat that:

1. Blocks you from sending specific keywords.
2. Restricts you from texting after a certain number of hours at work. For example, if David has already not left his work station for ten hours in a row, he’ll get a prompt to take a break.
3. Detects how you sound in your text and if better word choices can help you come across more professional. It’s easy to send an angry response in the middle of a heated argument.
4. Understands the emotions behind your text (like sarcasm) and blocks it
5. Delays outgoing chat, giving you time to cool off. Most likely, you won’t send the angry text at all
6. Works both ways: you get to interpret other person’s emotions better
7. Complies fully with the company and individual privacy laws and can be turned on/ off anytime

There’s still a long way; we start seeing tone detection in day to day communication.

Moreover, the accuracy of tone detection also depends on how good the software scanning it is. To a bot, “I don’t disagree with you,” may look very similar to “You are wrong.” The lesser the volume of data being analyzed, the more chances of misinterpretation.

If you wish to check whether a piece of text sounds good, go to Sentiment Analyzer and analyze it for free.

Food for thought

Imagine all this, absolutely free, in your personal chat tools like Whatsapp, Facebook Messenger, and WeChat.

How many arguments you would be able to avoid.

Sometimes, silence is the best answer. For other vulnerable situations, let the bots come to your rescue.