0:00:00.000,0:00:03.420
[MUSIC PLAYING]

0:00:03.420,0:00:03.920


0:00:03.920,0:00:06.290
I'm April Williams, and I[br]am an assistant professor

0:00:06.290,0:00:08.430
in the Department of[br]Communication Media.

0:00:08.430,0:00:11.520
I'm also faculty in the[br]Digital Studies Institute,

0:00:11.520,0:00:14.960
and I am an RCI fellow of[br]Research and Community Impact

0:00:14.960,0:00:17.570
fellow with the NCID.

0:00:17.570,0:00:22.260
The RCI fellowship is funding[br]my project on online dating.

0:00:22.260,0:00:24.180
The project is[br]called Swipe Rights,

0:00:24.180,0:00:26.390
and I am thinking[br]through advocating

0:00:26.390,0:00:28.760
for people and the[br]way that we use

0:00:28.760,0:00:30.980
apps and everyday technologies.

0:00:30.980,0:00:33.200
AI has a lot of[br]bias, whether that's

0:00:33.200,0:00:35.148
racial bias or gender bias.

0:00:35.148,0:00:37.190
So my project is really[br]trying to bring awareness

0:00:37.190,0:00:41.250
to that bias, especially where[br]dating apps are concerned.

0:00:41.250,0:00:44.310
The dating apps are only giving[br]us what they think we want,

0:00:44.310,0:00:45.660
and that's the problem.

0:00:45.660,0:00:47.960
All the social problems[br]that exist in our society

0:00:47.960,0:00:49.890
get amplified by algorithms.

0:00:49.890,0:00:52.290
They're not necessarily[br]created by them.

0:00:52.290,0:00:54.270
And we have to really[br]stop and think through,

0:00:54.270,0:00:56.970
do I want to use a[br]race filter on Hinge?

0:00:56.970,0:01:00.930
Do I want to put something in[br]my bio that is racially coded?

0:01:00.930,0:01:05.030
Do I want to train the algorithm[br]to look for things that

0:01:05.030,0:01:06.570
might be racially sensitive?

0:01:06.570,0:01:08.510
Most of the[br]responsibility for bias

0:01:08.510,0:01:12.020
that happens in dating[br]platforms falls on those

0:01:12.020,0:01:14.040
who are making the platforms.

0:01:14.040,0:01:17.750
There's a huge power imbalance[br]where those folks at the top

0:01:17.750,0:01:20.150
have most of the power,[br]and us individual users

0:01:20.150,0:01:21.810
have less of the power.

0:01:21.810,0:01:25.350
I want to empower users to[br]know what they're looking for,

0:01:25.350,0:01:28.140
to ask questions about where[br]their data is coming from,

0:01:28.140,0:01:30.750
what happens to their data when[br]they upload it to the apps,

0:01:30.750,0:01:33.500
and to really call out[br]app designers and platform

0:01:33.500,0:01:36.660
designers, to get them to be[br]more accountable to users,

0:01:36.660,0:01:39.500
to be more trustworthy and[br]to be more transparent.

0:01:39.500,0:01:41.360
I have a forthcoming[br]book which is

0:01:41.360,0:01:43.910
on online dating and the[br]algorithms that drive

0:01:43.910,0:01:45.540
the online dating process.

0:01:45.540,0:01:46.940
It's called Not My Type--

0:01:46.940,0:01:49.553
Automating Sexual[br]Racism in Online Dating.

0:01:49.553,0:01:51.470
And I think about it[br]both from the perspective

0:01:51.470,0:01:54.360
of online daters and[br]the things that we do,

0:01:54.360,0:01:56.540
so our individual swiping[br]behaviors, but then

0:01:56.540,0:01:59.810
also the algorithms[br]that program and drive

0:01:59.810,0:02:01.920
our swipe-based experiences.

0:02:01.920,0:02:05.390
I'm working on a guide to[br]accompany the book so that users

0:02:05.390,0:02:08.990
can know exactly what language[br]they need to use when they are

0:02:08.990,0:02:11.910
writing to their[br]representatives and saying,

0:02:11.910,0:02:14.550
hey, look, actually, big[br]tech needs to be regulated.

0:02:14.550,0:02:16.050
This is something[br]that I care about.

0:02:16.050,0:02:17.570
This is something[br]that I would vote for.

0:02:17.570,0:02:18.070
Right?

0:02:18.070,0:02:20.670
We vote on policy[br]with our dollars,

0:02:20.670,0:02:23.820
whether that is an in-app[br]purchase, downloading Tinder,

0:02:23.820,0:02:26.330
downloading Bumble, or[br]any of the other apps

0:02:26.330,0:02:27.360
that you might use.

0:02:27.360,0:02:30.840
Everything that we do is[br]really reinforcing demand.

0:02:30.840,0:02:33.230
So if we want dating[br]apps to sort of stop

0:02:33.230,0:02:36.180
and think about the racial bias[br]that's in their algorithms,

0:02:36.180,0:02:38.730
we have to encourage[br]them by saying,

0:02:38.730,0:02:40.110
no, this is unacceptable.

0:02:40.110,0:02:43.340
And the way that we do that[br]is by withholding our money

0:02:43.340,0:02:44.040
from them.

0:02:44.040,0:02:47.750
So ultimately, the goal would[br]be that dating platforms, dating

0:02:47.750,0:02:50.330
companies, would do[br]an internal audit

0:02:50.330,0:02:52.550
and really take accountability[br]for the algorithms

0:02:52.550,0:02:54.000
that they're currently using.

0:02:54.000,0:02:56.310
Hopefully, if there were[br]enough users that said,

0:02:56.310,0:02:59.130
I'm concerned about this,[br]I'm uncomfortable with this,

0:02:59.130,0:03:02.490
that would really indicate to[br]these programmers to really say,

0:03:02.490,0:03:04.130
OK, we need to[br]stop and reevaluate

0:03:04.130,0:03:06.680
the way that we collect[br]data, the way that we

0:03:06.680,0:03:10.610
use data, the way that we draw[br]inferences about our user base.

0:03:10.610,0:03:13.610
So with this grant, I'm really[br]pulling together several

0:03:13.610,0:03:15.350
of the leading[br]scholars in the field

0:03:15.350,0:03:17.490
and lots of other[br]practitioners as well

0:03:17.490,0:03:21.050
to think through what it[br]means for the individual user

0:03:21.050,0:03:24.590
to say to a company, hey, I'm[br]not comfortable with the way

0:03:24.590,0:03:25.920
that you're using my data.

0:03:25.920,0:03:28.810
Hey, we're not going[br]to take this anymore.

0:03:28.810,0:03:36.000