On October 8 and 10, Ukrainian drone strikes took out two Russian data centers. The result? Almost immediately, countless Russian bots disappeared from the internet.

What did this mean for social media users?

For a few days, many of the people I follow and interact with reported a real — and really positive — change in their online experiences.

In a bot-ridden internet, posts on certain topics would be swarmed with replies that made everything just feel worse. If you posted about politics or health or gender in any way, chances were good that “people” would pile on with all kinds of negativity, inflammatory comments, or misinformation.

In the days following the data center takedowns, person after person reported suddenly enjoying their time on social media.

“For the first time in a decade, I have been on SM [social media] all day without that infuriating feeling!!!”

“Are you on Twitter? It’s night and day since the bots disappeared. Feels like the original Twitter before that [expletive] Elon took it over.”

“I have been managing social media for over 20 years and never seen the internet to quickly revert to 2012 energy. It felt like what talking to people on the street feels like. The humanity is real. The inflated anger and interactions unchecked by META and all the others are fake.”

Influential people with large followings even started running “bot tests” where they would post an opinion that would ordinarily be attacked. Suddenly there were way fewer comments, and most of them were supportive.

What are bots doing?

Not all bots on social media are malicious. Some are informational and others are just advertising products.

But many social media bots have been designed to manipulate people and cause problems IRL, in the real world offline.

Dozens of academic studies have shown that bots do this by:

Spreading misinformation
Faking support for unpopular people and viewpoints
Influencing elections
Promoting social division
Supporting extreme and radical social positions

What’s more, real human beings are bad at figuring out who is a bot and who is a person — and when you spend more time on social media, you’re more likely to be duped by bots.

Thanks to amplification by social bots, bad information, anti-social positions, and low-quality ways of debating spread through society. And these low-quality ways of arguing can end up in your workplace.

Bots vs. humans

What’s going on when people (or people and bots) disagree?

You might not be able to identify who is and who isn’t a bot, but we can use the science of linguistic anthropology to figure out how to avoid common bot practices.

Let’s look at the components of a communication event (a social activity where language is playing an important role) and separate out bots from humans.

1. Participants

On social media, you’ll find a mixture of humans and bots.

In the workplace, discussions are (or at least should be) fully human — although we are seeing more AI-generated emails, posts, reports, etc.

2. Ends and goals

On social media, the goals of bots are anti-social. Interactions are meant to make you feel angrier and worse, to make your views more extreme, to make society feel more divided. And to keep you arguing. Endless, eternal arguing.

In the workplace, the goal of meetings, brainstorming sessions, feedback sessions, etc. is (in principle) to make things better. To identify problems or gaps or disagreements and figure out solutions. To move from debate to implementation.

(Yes, we all know people whose goals are just to get more power and control. But luckily for us, they are in the minority.)

Communication has lots of different functions. Besides sharing information and being a tool to reach consensus, language also creates, builds, and strengthens relationships. Linguists call this phatic communication, and it’s why we say things like “Hello” and “How’s it going?” and “What’s up?” In US English, “meaningless” phrases like these are saying, “We are two humans who are interacting and are not dangerous.” In many cultures, small talk serves the same purpose.

Regular conversations (if nothing goes terribly wrong) build trust, feelings of safety, and cohesion.

3. Channels of communication

Social bots mostly communicate through the written word. Sometimes posts are supplemented with images, links, and (often AI-generated) videos.

Social bots cannot communicate face to face. And this works in their favor when it comes to making things worse via language.

That’s because the farther you get from face-to-face communication, the easier it is to misunderstand someone. Just this week, I heard three separate stories about emails where the recipient was sure that the sender was a) mad at them; b) disappointed in them; c) someone who didn’t like them. Then, after an in-person conversation, the misunderstandings were cleared up. The sender a) wasn’t mad at them at all; b) thought they were doing good work; c) liked them just fine.

Text-only conversations also make it much easier to get mad and stay mad at someone. And it’s much easier to say something harsh or mean or aggressive on social media than it is directly to someone’s face.