Twitter is wondering whether Twitter is bad for society — and Jack Dorsey is starting new research to find out

Are the conversations that people have on Twitter “healthy”? The company plans to figure that out.

That’s according to a series of tweets sent Thursday by CEO Jack Dorsey, who first issued a pseudo apology for the kinds of aggressive and abusive content that Twitter has become known for.

“We have witnessed abuse, harassment, troll armies, manipulation through bots and human-coordination, misinformation campaigns, and increasingly divisive echo chambers,” Dorsey wrote. “We aren’t proud of how people have taken advantage of our service, or our inability to address it fast enough.”

So now Twitter has a new plan: It wants to measure the “health” of conversations on Twitter, something that Facebook is trying to do as well.

“If you want to improve something, you have to be able to measure it,” Dorsey continued. “The human body has a number of indicators of overall health, some very simple, like internal temperature. We know how to measure it, and we know some methods to bring it back in balance.”

It’s not clear how you measure the health of a human interaction (or a human-to-bot interaction), but Dorsey intends to find out. He mentioned Cortico, a nonprofit that proclaims a mission on its website to “foster a healthy public sphere,” as an organization that is inspiring Twitter’s work. Specifically, Dorsey mentioned four “indicators” that Twitter could use for measuring the health of conversations. Cortico described them like this on its website:

Twitter also issued an RFP — a request for proposal — to get more ideas from others outside the company.

“We simply can’t and don’t want to do this alone. So we’re seeking help by opening up an RFP process to cast the widest net possible for great ideas and implementations,” Dorsey tweeted.

For years, Twitter has served as a home for some of the nastier internet users — a place where online harassment and bullying and conflict thrived, thanks in part to Twitter’s commitment to free speech and its willingness to allow anonymous users.

The company has tried hard to clean that up in recent years. It rewrote its user rules and guidelines, and has started cracking down on bot accounts. It has even started to punish well-known users by removing their account verification, or banning them entirely.

Now it’s taking an interesting step to try and quantify the “health” of its own product.

It all seems to stem from the aftermath of the 2016 presidential election, in which we learned that Russian bots used Twitter to try and sow discord among voters. Ever since Donald Trump was elected and social networks were found to be somewhat responsible, tech companies have been grappling with their broader role in society.

Twitter’s move on Thursday feels like an olive branch of sorts for politicians in D.C. There has been speculation ever since the election that Congress may try and regulate social media companies that were caught sleeping at the wheel. This is a way for Twitter to show everyone that it’s proactively trying to improve discourse on its platform.

It’s one of the reasons Facebook is also trying to measure its impact on user health. Late last year, the social giant released a study that found that social media sometimes left people feeling crummy. The company then used that report as a reason to change its News Feed algorithm to start showing people more posts from friends and family, and less from publishers and businesses.

This feels like a similar moment from Twitter. The company is now asking an important question: What role do we play in the health of our users? It’s a question Twitter might not want the answer to.

This article originally appeared on Recode.net.