J. Nathan Matias: I’m Nathan Matias, founder of CivilServant. And we’re here today to talk about the Internet.

Firefighters extinguising a dumpster fire at Altus AFB

Now, in the room today are some tru­ly remark­able peo­ple. And I want you to know that in this slide there is more than just a dump­ster fire. There are also peo­ple in suits who are train­ing and ded­i­cat­ed to man­age that fire. And many of those peo­ple, when it comes to the Internet, are here in the room. You’re peo­ple who’ve been cre­at­ing and main­tain­ing our digitally‐connected spaces for decades. You’re com­mu­ni­ty mod­er­a­tors, researchers, facil­i­ta­tors, engi­neers, edu­ca­tors, cre­ators, bystanders, and pro­tec­tors. People who look beyond the dump­ster fire of what­ev­er pan­ic we have of the moment, rec­og­nize that there’s some­thing deeply impor­tant worth sav­ing, worth grow­ing, and step in.

And in fact, there are a lot of us. Even as com­pa­nies are expect­ed to do some­thing about online risks, far more of us take action than I think many of us rec­og­nize. According to a study by the Data & Society insti­tute, around 46% of American Internet users have tak­en some kind of action to inter­vene on the issue of online harass­ment. That’s almost 100 mil­lion peo­ple.

Despite this we do live in pes­simistic times, as soci­ety con­fronts the risks that come with liv­ing in a world that is inescapably dig­i­tal­ly con­nect­ed. And so, with many things com­pet­ing for our fears, it’s easy to become par­a­lyzed or uncer­tain about what to wor­ry about most and what to do about those prob­lems. Is dis­in­for­ma­tion real­ly unstop­pable? Are our our minds real­ly being hijacked? And when we wor­ry about our hopes for soci­ety slip­ping away, how can we mean­ing­ful­ly move from fear and uncer­tain­ty to do some­thing that we’re con­fi­dent can make a dif­fer­ence?

I imag­ine that’s how Wikipedians must’ve felt in 2007, after a year of remark­able growth that more than dou­bled the num­ber of active con­trib­u­tors to this incred­i­bly pre­cious glob­al resource.

Here’s the sto­ry as told by the researchers Aaron Halfaker, Stuart Geiger and their col­leagues about a moment of cri­sis that they expe­ri­enced.

So, eleven years ago the Wikipedia com­mu­ni­ty was just start­ing to reck­on with the scale and risks of van­dal­ism on the site as it became more and more relied on by peo­ple around the world. Wikipedia had been sued for false infor­ma­tion that was post­ed to this ency­clo­pe­dia that any­one can edit. And one per­son was even detained by Canadian bor­der agents after some­one added false infor­ma­tion to their bio, claim­ing that they were a ter­ror­ist.

Now, Wikipedians learned, like so many oth­er large online com­mu­ni­ties, that their col­lec­tive endeav­or had made them a basic part of mil­lions of people’s lives, with all of the accom­pa­ny­ing mess, con­flict, and risk. And like many plat­forms, the com­mu­ni­ty built social struc­tures and soft­ware to man­age those prob­lems. They cre­at­ed auto­mat­ed mod­er­a­tors like the neur­al net­work ClueBot which detects van­dal­ism, and mod­er­a­tion soft­ware that allowed peo­ple to train and over­see those bots. Thanks to these ini­tia­tives, Wikipedians were able to pro­tect their com­mu­ni­ty, cut­ting in half the amount of time that van­dal­ism stayed on Wikipedia’s pages. It’s now impos­si­ble to main­tain Wikipedia with­out these AI sys­tems.

Unfortunately, this amaz­ing­ly cre­ative, suc­cess­ful effort had some side‐effects. Soon after intro­duc­ing auto­mat­ed mod­er­a­tion, Wikipedia par­tic­i­pa­tion went from mas­sive growth into a slow decline. In a paper that appeared six years lat­er, Aaron, Stuart and their col­lab­o­ra­tors noticed that this dra­mat­ic change coin­cid­ed with the use of these scaled mod­er­a­tion sys­tems.

They were able to show that up to 30% of edits removed by these bots were of new­com­ers who they imag­ined and esti­mat­ed were actu­al­ly good‐faith con­trib­u­tors, peo­ple who might oth­er­wise have brought their unique per­spec­tives to this amaz­ing shared resource of human knowl­edge.

Screenshot of the WikiGalaxy project

You know, Wikipedia’s one of the great trea­sures of our times, and to save it Wikipedians orga­nized thou­sands of peo­ple in col­lab­o­ra­tion with AI sys­tems in what has got to be one of the largest vol­un­teer endeav­ors in his­to­ry to fight mis­in­for­ma­tion. Each of us relies on their endeav­or every sin­gle time we use Wikipedia. And yet, the long‐term behav­ioral side‐effects went unde­tect­ed for years.

Today’s com­mu­ni­ty research sum­mit is the first pub­lic event for CivilServant, a project that start­ed here at the MIT Media Lab and the Center for Civic Media, now incu­bat­ed by Global Voices, we orga­nize the pub­lic so that as we work togeth­er for a fair­er, safer, more under­stand­ing Internet, we can also dis­cov­er togeth­er what actu­al­ly works and spot any side‐effects of the pow­er­ful tech­nolo­gies and human process­es we bring to bear.

To open up today we’ll hear from Tarleton Gillespie and Latanya Sweeney, who will set the stage for think­ing about how people’s behav­ior is gov­erned online, and how it’s impor­tant to hold the tech­nol­o­gy and the com­pa­nies behind it account­able for the pow­er they have in soci­ety.

We’ll also hear from Ethan Zuckerman and Karrie Karahalios about ways that peo­ple orga­nize to under­stand sys­tems and cre­ate change, and some of the bar­ri­ers we may need to over­come as we try to make sense of the prob­lems, the injus­tices, the mess of a com­plex world, and try to cre­ate more under­stand­ing soci­eties online.

After a break, we’ll show you the mis­sion of CivilServant by shar­ing light­ning talks with com­mu­ni­ties that have already done research with us and orga­ni­za­tions that are part of our mis­sion.

We’ll then have time to hear from a num­ber of oth­er projects that are very much peers to what we do, and hear about the amaz­ing work that they’re doing.

When all of that ends, we’ll also have an hour for all of you to have some refresh­ments and meet each oth­er. Because as much as we are tak­ing the oppor­tu­ni­ty to share what we’ve done togeth­er, this room has an amaz­ing col­lec­tion of peo­ple with incred­i­ble per­spec­tives, wis­dom, and capa­bil­i­ties that will help us think about and shape the future of a fair­er, safer, more under­stand­ing Internet.

Today’s sum­mit is made pos­si­ble through fund­ing of the Ethics and Governance of AI Fund, the Knight Foundation, the MacArthur Foundation, sup­port from the Tow Center for Digital Journalism. Today’s event is pub­lic, and all of these talks will be filmed and put on YouTube. You should feel free to tweet about and share what you see on this stage. And we encour­age you, with people’s con­sent, to share what you heard, attribute what peo­ple tell you. But make sure to check with some­one before you share their face, their name, or their iden­ti­ty oth­er­wise. Some peo­ple in this room because of the nature of the work they do have cer­tain pref­er­ences for how they man­age their iden­ti­ty. So we ask that you respect people’s name badges and proac­tive­ly ask for con­sent before shar­ing.


Help Support Open Transcripts

If you found this useful or interesting, please consider supporting the project monthly at Patreon or once via Square Cash, or even just sharing the link. Thanks.