[00:00.000 --> 00:13.520]  ...talking to us and leading a conversation with all of you, encouraging you to solve the hard problem and maybe figure out how to finally make email encryption work.
[00:13.720 --> 00:15.420]  And wow, we... oh, wow, it works.
[00:15.960 --> 00:17.180]  It works for a second.
[00:18.160 --> 00:18.560]  Momentarily.
[00:18.740 --> 00:19.940]  Ignore everything on the screen.
[00:20.120 --> 00:20.960]  That's not her email.
[00:22.480 --> 00:23.360]  Okay, here we go.
[00:23.540 --> 00:24.860]  Okay, so Gus, it's all yours.
[00:25.040 --> 00:25.240]  Okay.
[00:25.880 --> 00:29.820]  I hope you don't mind if I... I'm gonna probably end up sitting down for most of this and do leave the house lights on.
[00:29.820 --> 00:31.540]  And this will be a very interactive presentation.
[00:32.740 --> 00:35.140]  So if you're in the back, come forward a little bit.
[00:35.200 --> 00:35.640]  Everybody come up.
[00:36.100 --> 00:37.580]  Actually, we're quite filled in in front.
[00:37.660 --> 00:38.080]  This should be good.
[00:38.240 --> 00:39.320]  So hi, I'm Gus Andrews.
[00:41.280 --> 00:44.640]  And I'm sort of running on not that many cylinders today, so bear with me.
[00:49.830 --> 00:50.230]  We good?
[00:50.430 --> 00:50.910]  Okay, we're good.
[00:52.190 --> 00:53.450]  I promise I won't drop any mics.
[00:53.750 --> 00:55.990]  It'll look like I'm about to drop some mics, but I'm not gonna drop some mics.
[00:56.150 --> 00:59.150]  So first I need to note that I'm not speaking on behalf of my job.
[00:59.170 --> 01:00.410]  These are all my personal opinions.
[01:00.590 --> 01:00.730]  Good.
[01:00.730 --> 01:09.270]  So, esteemed geeks, today I'm going to make the argument to you that your belief in science is wrong.
[01:09.870 --> 01:14.110]  And what's more, it may be responsible for why Johnny can't encrypt.
[01:15.010 --> 01:15.790]  All right?
[01:17.510 --> 01:19.430]  I'm gonna illustrate my points using this.
[01:19.590 --> 01:20.310]  Anybody know what this?
[01:20.550 --> 01:22.030]  Somebody, what is this?
[01:23.290 --> 01:25.390]  SKCD is our beloved geek comic.
[01:25.550 --> 01:29.370]  This is his, Randall Monroe's Fields Arranged by Purity.
[01:29.970 --> 01:36.810]  So I'm gonna illustrate my points using this and also illustrate my points with my ex-boyfriend who two hopes ago left me to go into the Peace Corps.
[01:36.810 --> 01:37.670]  Right?
[01:37.870 --> 01:39.150]  As you see him here.
[01:40.270 --> 01:41.850]  He went... Boo!
[01:42.470 --> 01:44.710]  I think maybe I'm biasing the audience already.
[01:46.210 --> 01:47.870]  That's not his actual face.
[01:49.070 --> 01:52.670]  He went to an African country I'll refer to as... Hello.
[01:53.130 --> 01:53.630]  I'm sorry.
[01:53.950 --> 01:55.590]  Oh, I'm so sorry, time machine.
[01:55.710 --> 01:55.970]  F*ck you.
[01:57.610 --> 01:58.290]  Pardon me.
[01:58.870 --> 01:59.470]  I know.
[01:59.610 --> 01:59.990]  I'm sorry.
[01:59.990 --> 02:02.950]  If there's any kids in the room, you may want to cover their ears.
[02:04.430 --> 02:04.750]  Right.
[02:04.990 --> 02:07.390]  So, CompSci, this is my...
[02:07.390 --> 02:09.030]  He called himself CompSci.
[02:09.150 --> 02:10.790]  This was a matter of family pride for him.
[02:10.890 --> 02:11.890]  He was a computer scientist.
[02:12.070 --> 02:13.410]  His father was a computer scientist.
[02:13.550 --> 02:16.010]  So every now and again, he'd just say, well, CompSci, you know.
[02:16.610 --> 02:17.950]  He refers to himself as CompSci.
[02:18.830 --> 02:24.710]  So CompSci went to an African country that I'll refer to as Fulani, to obscure where it was because otherwise you'll find his blog.
[02:24.850 --> 02:25.210]  Don't do that.
[02:26.710 --> 02:27.810]  He's not in the room, is he?
[02:28.850 --> 02:30.170]  I really hope he's not in the room.
[02:30.350 --> 02:31.730]  Anyway, Ethan, if you're here, you might want to leave.
[02:33.050 --> 02:34.170]  Todd, if you're here, don't tell...
[02:34.170 --> 02:35.410]  Is my boss in the room for that matter?
[02:35.470 --> 02:35.710]  Never mind.
[02:35.830 --> 02:35.930]  Sorry.
[02:36.810 --> 02:38.810]  Anyway, so CompSci was basically his way of identifying.
[02:39.570 --> 02:43.050]  This is his way of identifying and his way of doing things was a CompSci way of doing things.
[02:43.930 --> 02:47.690]  If he identified with coding so much that he called himself CompSci, where does he belong on here?
[02:47.790 --> 02:49.710]  I'm going to ask for hands rather than call outs.
[02:50.170 --> 02:52.210]  Where do you belong on this chart if he's in CompSci?
[02:55.820 --> 02:56.800]  There will not be a quiz.
[02:56.900 --> 02:57.060]  Yes.
[02:58.040 --> 02:59.700]  A little bit ahead of the physicist.
[03:00.200 --> 03:01.320]  A little bit ahead of the physicist.
[03:01.480 --> 03:03.620]  Actually, that is exactly where I have him placed, right?
[03:03.860 --> 03:06.600]  So it's math-y stuff, right?
[03:06.680 --> 03:09.700]  So it's getting close to pure, but it's also applied, right?
[03:09.760 --> 03:10.780]  So that's not very pure.
[03:10.880 --> 03:11.420]  It's applied.
[03:12.040 --> 03:14.120]  Oh, hey, I forgot to turn off my phone.
[03:16.760 --> 03:17.700]  Oh, that's awesome.
[03:18.600 --> 03:20.440]  That was actually my boss, too.
[03:20.640 --> 03:20.900]  That's great.
[03:22.340 --> 03:23.360]  Greatest talk ever.
[03:23.640 --> 03:25.260]  Great talk or greatest talk.
[03:25.760 --> 03:26.620]  Sorry about this, folks.
[03:27.420 --> 03:30.720]  Right, so yeah, so we got him settled on this chart.
[03:30.820 --> 03:34.120]  Is there anything over to the right of this chart, further past mathematicians?
[03:35.160 --> 03:35.320]  What?
[03:36.500 --> 03:36.900]  Philosophers.
[03:37.000 --> 03:38.180]  Some people might say philosophers.
[03:38.460 --> 03:39.700]  I don't think Randall Munro agrees with you.
[03:39.940 --> 03:40.840]  How about to the left?
[03:40.960 --> 03:42.220]  What is over to the left on this chart?
[03:45.340 --> 03:47.180]  We also heard philosophers again, right?
[03:48.080 --> 03:48.820]  Who else?
[03:48.980 --> 03:49.220]  What else?
[03:51.680 --> 03:52.080]  Artists!
[03:52.740 --> 03:54.440]  Who knows where the hell they are?
[03:54.680 --> 03:57.620]  It's just somewhere off in the... you tip the chart, and they're all the way down at the...
[03:57.620 --> 04:05.100]  Down at the end, this is from his other strip, which is Every Major is Terrible, which is absolutely one of my favorites of his strips, saying that every major in academia is terrible.
[04:05.300 --> 04:06.480]  So yeah, poetry's over there.
[04:06.880 --> 04:11.580]  Why anyone who wants a job would study lit's a mystery unless their only choice was something like art history?
[04:12.220 --> 04:15.680]  And so, yeah, he thinks that philosophy is just Matt Sanzerger's sense and practicality, right?
[04:15.840 --> 04:19.240]  So anyway, you go further left on this chart, you get maybe anthropology, right?
[04:19.860 --> 04:23.060]  Philosophy, maybe, depending on how you think of things, history, literature, poetry.
[04:24.680 --> 04:30.420]  Who here has... show of hands, who here has sensed that this is a hierarchy in the world of what's important to study?
[04:30.560 --> 04:31.760]  Who's had that sense at some point?
[04:32.920 --> 04:33.560]  Be honest.
[04:33.800 --> 04:34.140]  There we go.
[04:34.220 --> 04:34.520]  Yeah, right?
[04:34.620 --> 04:37.060]  So this is something that we're all pretty familiar with in a way, right?
[04:37.780 --> 04:39.180]  Where did you get that sense?
[04:40.040 --> 04:41.420]  This is not a rhetorical question.
[04:41.580 --> 04:42.180]  Anybody have any ideas?
[04:42.260 --> 04:43.120]  Where did you get that sense?
[04:43.360 --> 04:43.800]  High school.
[04:43.920 --> 04:44.920]  High school, okay.
[04:45.160 --> 04:45.420]  What?
[04:45.980 --> 04:46.340]  Salaries.
[04:47.520 --> 04:47.880]  Salaries.
[04:48.420 --> 04:49.320]  Oh, yes.
[04:49.860 --> 04:53.580]  Yes, your boss likes you better if you do some math, especially applied math.
[04:53.720 --> 04:53.880]  It's great.
[04:54.020 --> 04:54.560]  So what else?
[04:54.880 --> 04:55.240]  Magazines.
[04:55.760 --> 04:56.120]  Magazines.
[04:56.280 --> 04:57.340]  Okay, media, generally, right?
[04:57.420 --> 04:59.160]  There might be some media reasons why, right?
[04:59.480 --> 04:59.560]  So...
[04:59.560 --> 04:59.700]  It's not good.
[05:01.960 --> 05:03.780]  Okay, so now we know where the trolls are sitting.
[05:03.780 --> 05:04.180]  All right.
[05:05.540 --> 05:08.840]  Okay, so yeah, so purity is sort of an odd way to put this, right?
[05:09.160 --> 05:09.480]  Purity...
[05:09.480 --> 05:10.280]  I mean, so we've heard other things.
[05:10.400 --> 05:12.080]  We've heard like social hierarchies in schools.
[05:12.100 --> 05:14.200]  We've heard what you get paid, right?
[05:14.340 --> 05:15.300]  So those are other ways of doing this.
[05:15.380 --> 05:16.920]  Purity is a bit of an odd one, but it's...
[05:16.920 --> 05:20.060]  Purity is maybe the intellectual way of sort of viewing this particular chart.
[05:21.580 --> 05:26.860]  If you weren't going to say purity, how else would you describe who's way over on the right-hand side versus who's all the way over on the left?
[05:27.000 --> 05:28.160]  What words would you use?
[05:28.840 --> 05:29.160]  Pedantism.
[05:30.220 --> 05:30.540]  Pedantism.
[05:30.540 --> 05:30.960]  Who else?
[05:31.500 --> 05:31.860]  Abstraction.
[05:32.540 --> 05:32.900]  Abstraction.
[05:33.120 --> 05:33.220]  Okay.
[05:34.180 --> 05:34.540]  Asperger's.
[05:35.340 --> 05:35.700]  Asperger's.
[05:35.800 --> 05:36.460]  That's actually probably pretty good.
[05:36.780 --> 05:37.820]  So going into...
[05:38.820 --> 05:39.700]  We've got to be fair though.
[05:40.060 --> 05:42.140]  If we're going into abstraction, like what...
[05:42.140 --> 05:46.240]  If people are more abstract, if they're able to handle more abstraction, what would you say about them?
[05:47.100 --> 05:48.000]  They can generalize.
[05:48.380 --> 05:49.080]  They can generalize.
[05:49.160 --> 05:49.420]  What else?
[05:50.640 --> 05:51.700]  Right brain, left brain.
[05:51.780 --> 05:53.520]  So we're talking about brains in some way.
[05:53.640 --> 05:54.740]  Maybe we're talking about smarts.
[05:54.740 --> 05:55.520]  Who would say maybe smart?
[05:56.300 --> 05:57.480]  Smarter people over to the right.
[05:58.000 --> 05:58.460]  Ah, yes.
[05:58.620 --> 05:58.780]  Okay.
[05:58.780 --> 05:59.200]  No?
[05:59.720 --> 06:00.180]  No, no.
[06:00.300 --> 06:02.180]  I've seen some heads nodding and I'm seeing some heads shaking.
[06:02.300 --> 06:02.360]  No.
[06:02.480 --> 06:03.080]  So that's good to know.
[06:03.280 --> 06:03.360]  Okay.
[06:05.820 --> 06:07.320]  How about what is useful?
[06:08.140 --> 06:09.520]  What is practically implied and useful?
[06:09.640 --> 06:11.040]  So maybe over to the right it's more useful, maybe?
[06:11.220 --> 06:11.400]  Maybe?
[06:11.540 --> 06:11.740]  I don't know.
[06:11.860 --> 06:12.040]  Math?
[06:12.140 --> 06:12.400]  Who knows?
[06:13.280 --> 06:13.600]  Right?
[06:14.700 --> 06:15.580]  How do we know?
[06:15.960 --> 06:19.520]  Once again, how do we know what the values are on each of these parts?
[06:24.260 --> 06:24.580]  Market.
[06:26.000 --> 06:29.560]  So once again, maybe going back to school, maybe going back to social structures, things like that.
[06:29.820 --> 06:30.140]  So, okay.
[06:31.080 --> 06:31.560]  Media.
[06:33.640 --> 06:35.120]  School should be useful.
[06:35.300 --> 06:35.320]  Right?
[06:36.040 --> 06:36.300]  Okay.
[06:36.540 --> 06:45.460]  So moving onward, this talk is going to be about why it's a huge problem to value math, science, and technical kinds of knowledge at the expense of other kinds of knowledge.
[06:45.720 --> 06:51.180]  Why it's a problem to treat those problems as the most challenging, interesting, and important problems, right?
[06:51.180 --> 06:55.780]  Are physics and math and science the most important problems, the most challenging problems?
[06:55.840 --> 06:57.500]  I think a lot of people here would find them challenging, right?
[06:59.860 --> 07:08.380]  I'm problematizing this because it doesn't always go well when you give free reign to someone whose only credentials were that they were a tech engineer at 17 and then some sort of genius physics major.
[07:09.700 --> 07:11.940]  For example, Second Life, where I worked briefly.
[07:12.280 --> 07:13.240]  I don't know if you're familiar.
[07:13.240 --> 07:22.960]  Phillip Rosedale was a physics major and also at 17 a tech entrepreneur.
[07:23.180 --> 07:26.600]  So he says, in the virtual world, things are easier and people are more honest.
[07:26.780 --> 07:30.560]  This is spoken by somebody who's not at all inquisitive about what people are actually doing, right?
[07:30.820 --> 07:38.520]  So Phillip hired... he had this sort of like vaunted, much talked about in the media way of running his company that was like, yeah, it's going to be like a free market.
[07:38.680 --> 07:44.620]  And like within the company, whoever does the coolest, smartest thing that's the most valuable will be rewarded by other people in the company.
[07:44.720 --> 07:46.800]  It was a very structuralist, flat hierarchy.
[07:47.660 --> 07:52.580]  And so he hired a whole company full of math, science, and code geniuses and basically let them do what they wanted.
[07:52.980 --> 07:59.160]  The result would send the most patient gamer crying to his parents' living room to watch Everybody Loves Raymond, right?
[07:59.780 --> 08:01.120]  Like forget the hard problems.
[08:01.220 --> 08:02.120]  They'd be like, I'm done with this.
[08:02.240 --> 08:03.600]  I'm just done with the hard problems.
[08:03.720 --> 08:04.180]  Oh my God.
[08:04.600 --> 08:04.760]  Right?
[08:04.940 --> 08:11.280]  So, I mean, you also ended up with high-end graphics that were awesome until you used them on like a netbook, which people regularly did, right?
[08:11.280 --> 08:13.620]  Because they were like, yeah, we're going to make really shiny graphics.
[08:13.780 --> 08:16.960]  And then they rendered absolutely poorly on what was not top of the line software, right?
[08:17.280 --> 08:19.040]  I mean, there's naked people in this frame.
[08:19.180 --> 08:20.880]  Did you even notice that there were naked people in this frame?
[08:20.960 --> 08:23.540]  You have to wait five minutes to find out that they're there, right?
[08:23.940 --> 08:25.520]  Second Life, great success.
[08:25.780 --> 08:27.160]  Okay, stop him.
[08:27.300 --> 08:32.180]  He's about to start another virtual world startup shortly.
[08:32.400 --> 08:33.640]  So stop him before he kills again.
[08:33.740 --> 08:34.940]  Anyway, I worked at Second Life.
[08:35.800 --> 08:36.800]  I have some biases.
[08:36.960 --> 08:37.620]  I worked at Second Life.
[08:37.700 --> 08:38.960]  It was actually a really great place to work.
[08:39.020 --> 08:39.700]  It was a lot of fun.
[08:39.700 --> 08:43.500]  There were some serious problems in, you know, what was being prioritized.
[08:43.680 --> 08:47.220]  But the main thing I didn't understand is why everybody fawned all over Philip.
[08:47.400 --> 08:49.240]  They're like, oh, yeah, he's such a great market leader.
[08:49.340 --> 08:50.080]  He's a tech whiz kid.
[08:50.140 --> 08:50.760]  He studied physics.
[08:50.980 --> 08:57.980]  I think mostly they fawned over him because he was like Mark Hamill, only with hypnotic eyes that would drill through you while he was talking, right?
[08:58.180 --> 09:04.340]  So anyway, this kind of fawning over geniuses for their brilliant tech solutions is a constant these days, right?
[09:04.400 --> 09:05.060]  You've got Bre Pettis.
[09:05.060 --> 09:06.540]  You've got Nicholas Negroponte.
[09:07.360 --> 09:13.420]  Negroponte ran multiple failed laptop programs before OLPC, the one laptop per child program.
[09:13.620 --> 09:16.360]  And yet we still were like, oh, yeah, he's that guy who does that thing.
[09:16.460 --> 09:19.660]  We should give him more laptops and send him to Fulani, apparently.
[09:20.100 --> 09:23.060]  Why does everybody at MakerBot do what Bre Pettis says?
[09:24.200 --> 09:24.800]  I don't know.
[09:25.220 --> 09:28.260]  He's actually, there's a movie coming out about, and we were going to screen it.
[09:28.260 --> 09:35.000]  We didn't end up getting it to screen, but it's worth checking out the MakerBot movie, because that'll give you a background into what's going on in this brilliant tech company.
[09:36.200 --> 09:42.440]  So what if math and science don't make you smart because they're better than other fields in some way, that they are more rigorous, more abstract?
[09:42.700 --> 09:49.580]  What if we're viewing it backward and math and science are valued most highly because they produce what capital wants most, right?
[09:51.620 --> 09:56.020]  Standardization, automated production, and a constant stream of new products.
[09:56.020 --> 09:56.120]  Yes.
[09:57.940 --> 09:59.120]  Anyway, back to CompSci.
[09:59.240 --> 10:02.520]  CompSci enrolled in Peace Corps, and I wanted to support him as much as I could.
[10:02.640 --> 10:09.660]  You know, I just finished my degree in Anthropology and Education and Technology, and I thought I could maybe help him because he's going to a country and he just didn't understand the people there.
[10:10.340 --> 10:17.220]  But his students said things and did things that he didn't understand, like, well, my computer doesn't have open all programs, right?
[10:17.900 --> 10:19.800]  So at one point, he wrote a blog post about this.
[10:20.300 --> 10:25.040]  He said, they're just not able to reason through basic things like finding where all programs are.
[10:25.500 --> 10:27.860]  They don't look in any of the right places on the screen.
[10:29.120 --> 10:32.440]  And I said, okay, well, what's going on?
[10:32.500 --> 10:37.040]  He said, like, I've got this lab where I've got Windows XP machines and Ubuntu machines and Xubuntu machines.
[10:37.080 --> 10:39.740]  And I said, are you switching them around from computer to computer?
[10:40.100 --> 10:40.920]  He said, yeah, of course.
[10:40.980 --> 10:41.900]  I want them to learn different systems.
[10:42.040 --> 10:44.940]  I'm like, that was the point at which I was like, well, there's your problem.
[10:45.140 --> 10:50.440]  Like, I was looking under the hood in this car and that didn't go over particularly well because he was smart and he did not want to be unsmart suddenly.
[10:52.960 --> 11:00.840]  So, yeah, what's your hypothesis about why his computer was having, what was, sorry, now, why is, why is students, why were students having problems finding all programs?
[11:03.120 --> 11:05.100]  The directions were written by the march.
[11:06.220 --> 11:07.020]  The, sorry?
[11:07.580 --> 11:09.080]  They were learning one system.
[11:09.300 --> 11:09.420]  Yep.
[11:09.660 --> 11:11.220]  And then being exposed to something that is completely different.
[11:11.500 --> 11:12.720]  Something completely different, right?
[11:12.800 --> 11:16.980]  And this is not, like, it's, I asked him, you know, do your students know what an operating system is?
[11:16.980 --> 11:20.520]  And he said, no, no, they know that there are some computers in English and some computers in French.
[11:20.540 --> 11:26.220]  So they didn't even have a, a means of, of conceptualizing why suddenly everything's moved around on the screen, right?
[11:26.320 --> 11:29.020]  So to them, it could just look like technology was completely broken, right?
[11:29.660 --> 11:31.820]  None of these are things that a human being is more knowing, right?
[11:31.900 --> 11:33.680]  We have all had tremendous amounts of practice.
[11:33.740 --> 11:36.660]  And these are folks who maybe hadn't seen a computer as often as we had.
[11:36.820 --> 11:38.500]  So they had no means of conceptualizing this.
[11:39.940 --> 11:41.820]  They also said other things that he didn't understand.
[11:41.820 --> 11:46.480]  And he was sort of struggling to, you know, to sort of figure out what, why they were saying it.
[11:47.000 --> 11:53.200]  This was something that his student said to him, um, when they saw him swishing his foot in his sandal, uh, in a puddle of water.
[11:53.400 --> 11:55.900]  And he was like, that's totally illogical and irrational.
[11:56.120 --> 11:56.980]  Why did they say that?
[11:57.060 --> 11:57.960]  Does anybody have a hypothesis?
[11:58.140 --> 11:59.180]  And I'd like to see hands, once again.
[11:59.360 --> 12:01.460]  Why, why, what was your, his, his student thinking?
[12:01.640 --> 12:01.780]  Yes.
[12:05.000 --> 12:05.620]  Exactly, right.
[12:05.660 --> 12:06.600]  It could have been a muddy puddle.
[12:06.680 --> 12:07.260]  You can't see the bottom.
[12:07.340 --> 12:08.940]  I thought it was a perfectly rational thing to say.
[12:09.020 --> 12:10.700]  And he was like, these people have no logic.
[12:10.900 --> 12:14.840]  So I was, anyway, so he would go around and say, I want these people to see logic and reason.
[12:14.840 --> 12:18.180]  That's not a direct quote, but he came very close to saying that, um, right?
[12:18.540 --> 12:20.220]  He never figured out these logics.
[12:20.280 --> 12:23.100]  And I said to him, look, they have their own cultural logic.
[12:23.260 --> 12:25.960]  They have a logic that they agree upon with the people around them, right?
[12:26.040 --> 12:34.980]  Like there are things that make sense to them because they've, you know, sort of come to it together by observing the world around them, by observing the operating system that maybe suddenly changed, right?
[12:35.100 --> 12:37.800]  By observing the puddles that are in their area, whatever, the water.
[12:38.760 --> 12:43.420]  So he says to me, culture is a hand wavy word you use when you don't understand what's going on.
[12:48.660 --> 12:50.080]  I've demonized him too quickly.
[12:50.220 --> 12:51.900]  Everybody's like, no, we're not on his side.
[12:52.040 --> 12:52.800]  We're on your side.
[12:52.960 --> 12:55.780]  So yeah, that was the point at which I was like, okay.
[12:56.640 --> 12:58.240]  But you know, you can't exactly blame him.
[12:58.240 --> 13:00.020]  He was just going off this chart, right?
[13:00.560 --> 13:00.920]  Culture.
[13:01.180 --> 13:02.900]  Where is studies culture on here?
[13:04.120 --> 13:04.800]  To the left.
[13:06.380 --> 13:06.780]  What?
[13:06.880 --> 13:07.280]  Gestures.
[13:07.800 --> 13:10.620]  Where to the, where, where, which fields?
[13:10.780 --> 13:11.440]  What are we talking about?
[13:13.700 --> 13:14.100]  Anthropology.
[13:14.280 --> 13:14.380]  So right.
[13:14.540 --> 13:16.360]  So anthropology we established earlier is a little bit further off.
[13:16.620 --> 13:17.680]  Sociology and psych, right?
[13:17.880 --> 13:17.960]  Okay.
[13:22.470 --> 13:25.710]  So it's not even, I mean, like really culture itself.
[13:25.890 --> 13:27.330]  Anthropology is really more the study of culture.
[13:27.450 --> 13:29.630]  Sociology is the study of social structures mostly.
[13:29.630 --> 13:38.430]  So, um, and so, uh, what can we say about that way of, um, no, we've established, right?
[13:38.530 --> 13:41.250]  So he, he believes that basically it's less pure, it's less rigorous, right?
[13:41.430 --> 13:42.590]  So, yeah.
[13:42.610 --> 13:43.750]  So here are the problems with what this guy said.
[13:43.850 --> 13:49.870]  First of all, he basically was saying other cultural logics don't matter, um, because they're not based on empirical evidence.
[13:49.950 --> 13:55.630]  That was the thing that he'd say is these people are not rationally reasoning based on what's around them, which I think was bogus.
[13:56.270 --> 14:02.190]  And he said other, he was basically also saying other ways of doing research to understand the world don't matter because they're impure.
[14:02.290 --> 14:08.170]  But the real problem with what this asshole said was everything my girlfriend has done for the past seven years is worthless, even though I don't understand it.
[14:08.230 --> 14:13.750]  So I was like, sayonara, you go stay in Africa for another couple of years, just, just stay there.
[14:14.430 --> 14:15.510]  He came back eventually.
[14:17.130 --> 14:22.190]  So if you want to know why a group of people are acting a certain way, how do you ask that question?
[14:22.930 --> 14:24.590]  Will you do science to it?
[14:26.030 --> 14:29.650]  How will you scientifically know how humans are behaving, right?
[14:29.910 --> 14:32.110]  Let's go back to the scope of our question and where it fits on here.
[14:34.250 --> 14:37.230]  If math is more pure, why don't we do math to everything?
[14:37.390 --> 14:41.170]  Think on this for a second and then give me some hands, because this is sort of an answer that would take some work.
[14:41.510 --> 14:45.930]  If math is more pure and more rigorous, why don't we do math to absolutely everything?
[14:46.070 --> 14:47.530]  Or should we do math to absolutely everything?
[14:54.510 --> 14:59.190]  They won't always react the same way to the same set of stimuli.
[14:59.330 --> 15:02.190]  So we're still thinking of them in sort of a scientific term if we're saying stimuli, right?
[15:02.910 --> 15:05.090]  I would say we don't understand the components of the system.
[15:05.090 --> 15:07.090]  Sorry, if I haven't called on you, please don't.
[15:07.210 --> 15:07.590]  Sorry, go ahead.
[15:07.770 --> 15:12.410]  We don't understand the components of the system and therefore we can't understand it mathematically yet.
[15:12.510 --> 15:17.470]  Ah, we don't understand the components of the system and therefore we can't understand it mathematically yet.
[15:17.510 --> 15:18.130]  That's well phrased.
[15:18.130 --> 15:18.630]  Yes, thank you.
[15:19.310 --> 15:21.630]  And furthest in the back hand back there.
[15:21.990 --> 15:22.390]  Yes?
[15:23.190 --> 15:27.470]  So we do, but we use applications, and we use these applications in the wrong way.
[15:27.630 --> 15:28.750]  So we're just using them in the right way.
[15:29.130 --> 15:29.830]  Applications of what?
[15:30.110 --> 15:32.350]  Of mathematics, abstract mathematics.
[15:32.610 --> 15:34.470]  Okay, so we do use mathematics to understand.
[15:34.570 --> 15:37.710]  So like what kinds of, what kind of way are we using mathematics to understand human behavior, would you say?
[15:39.250 --> 15:46.430]  Typical analysis of different ways of organizations exist in countries, communicate war, things like that.
[15:46.430 --> 15:52.090]  Okay, so we're sort of looking at the big social institutions and sort of what kinds of, how they operate, stuff like that.
[15:52.290 --> 15:53.430]  I've heard from you quite a bit.
[15:53.610 --> 15:54.730]  Gentleman over there standing up.
[15:56.610 --> 15:59.110]  Any mathematical description is going to
[16:02.450 --> 16:03.770]  model mathematically.
[16:04.190 --> 16:12.710]  When it comes to human relationships, the parts that are most important are the least amenable to the application of math.
[16:13.030 --> 16:14.610]  That's also very well said, right?
[16:14.730 --> 16:16.530]  So there are things that we do do math to, right?
[16:16.530 --> 16:23.230]  And we're going to sort of walk through which parts of social behavior, you know, people, we sort of are amenable to math and other parts aren't.
[16:23.370 --> 16:24.510]  And so we'll sort of explore those.
[16:24.690 --> 16:26.030]  I want to make sure that I keep moving here.
[16:26.110 --> 16:29.050]  I mean, another question is why don't we do biology to everything that humans do?
[16:29.150 --> 16:31.050]  Or should we do biology to everything that humans do?
[16:32.050 --> 16:32.710]  Way in the back.
[16:32.810 --> 16:32.870]  Yeah.
[16:38.060 --> 16:39.660]  What doesn't really take randomness into account?
[16:51.880 --> 16:55.060]  Math is particularly not good, I think, at understanding external environment, right?
[16:55.160 --> 16:55.200]  Yeah.
[16:55.320 --> 16:59.240]  So external environment, if we were studying a human's external environment, which field would we look at here?
[17:01.560 --> 17:04.060]  But seriously, which fields would you talk about for external environment here?
[17:05.600 --> 17:07.340]  Sociology, other parts of the external environment?
[17:08.500 --> 17:09.080]  Biology, right?
[17:09.180 --> 17:09.240]  Yeah.
[17:09.380 --> 17:12.740]  So I mean, there are multiple factors that should maybe be studied in different ways, right?
[17:12.740 --> 17:15.760]  And the whole point about randomness, I think, is also quite good.
[17:18.220 --> 17:20.520]  Math... chaos theory is good for randomness, right?
[17:20.640 --> 17:23.300]  So that ends up... it's kind of all over the place, actually.
[17:23.460 --> 17:25.880]  Any field here feels like it could have chaos theory, ultimately.
[17:26.140 --> 17:29.060]  But one of the things that strikes me is a lot of the time...
[17:29.660 --> 17:32.600]  I mean, you have a massive environment to take into account.
[17:32.780 --> 17:37.320]  So you're sort of trying... you're always... when you're doing research, you're always trying to sort of like sort out certain variables.
[17:38.540 --> 17:44.260]  So there are... yeah, there will be things that will pop up that appear to be noise.
[17:45.060 --> 17:47.560]  There are things that pop up that are more like history, right?
[17:47.720 --> 17:48.940]  There are... why did this happen?
[17:48.980 --> 17:50.340]  You have to look back in time.
[17:50.660 --> 17:53.300]  And math isn't necessarily good at expressing time.
[17:53.580 --> 17:55.700]  Biology is not necessarily good at expressing time.
[17:56.340 --> 17:59.620]  Depending on... it's like depending on how you want to document and what you want to document.
[17:59.760 --> 18:00.220]  Let's keep moving.
[18:00.360 --> 18:01.440]  I just... I feel like we're bogging down on this.
[18:01.500 --> 18:02.620]  But this is also a very good discussion.
[18:02.780 --> 18:07.560]  Like this is sort of... I'm sort of trying to come to epistemologically, which is the study of knowledge.
[18:07.720 --> 18:11.260]  Knowledge-wise, why is it that we come to the reasons that we do?
[18:11.800 --> 18:14.080]  Let's work on a more specific question, right?
[18:14.200 --> 18:16.980]  So that we can sort of apply the reasoning behind things here.
[18:17.280 --> 18:21.240]  Say we wanted to answer the question, why can't Johnny encrypt by doing research?
[18:21.700 --> 18:29.360]  How could we do a quantitative, a mathematical study about why people are and aren't encrypting?
[18:29.740 --> 18:30.800]  What could we do?
[18:31.660 --> 18:32.020]  Why?
[18:32.880 --> 18:36.540]  What could we... how could we do a mathematical study about why people aren't encrypting?
[18:36.580 --> 18:37.680]  You look like you might have an answer.
[18:38.100 --> 18:39.220]  In the blue, right there.
[18:39.360 --> 18:39.960]  No, you're just stretching.
[18:42.840 --> 18:49.400]  Yeah, I think surveys before and after as you do compile the results and find out what the most common results are.
[18:49.440 --> 18:49.640]  Okay.
[18:50.080 --> 18:52.280]  So if we're doing surveys...
[18:52.280 --> 18:54.620]  So surveys, do you mean like people writing things down for us?
[18:54.780 --> 18:54.940]  Right.
[18:54.940 --> 18:55.340]  Answering questions?
[18:55.520 --> 18:56.900]  So that would actually be qualitative, right?
[18:56.900 --> 19:00.640]  I mean, so you're getting... to some extent you're getting a word response.
[19:00.780 --> 19:03.920]  How do we do like sheer pure math to Glenn Greenwald and why he's not doing it?
[19:04.000 --> 19:04.120]  Yes.
[19:04.360 --> 19:06.640]  Count the number of people that have access to encryption programs.
[19:06.760 --> 19:09.180]  Count the number of people that have access to encryption programs.
[19:09.300 --> 19:10.080]  That's an interesting one, right?
[19:10.220 --> 19:11.760]  So how would we know if they have access?
[19:14.140 --> 19:15.160]  Do they have Internet access?
[19:15.680 --> 19:18.980]  Does that also tell us whether they're aware of encryption programs?
[19:19.780 --> 19:20.500]  No, right?
[19:20.600 --> 19:22.060]  So we've already missed something there.
[19:22.560 --> 19:26.340]  So we have to not just look at do they have...
[19:26.340 --> 19:29.140]  Is there the possibility that they could get it because they have the Internet?
[19:29.380 --> 19:31.920]  But we also have to know are they aware?
[19:32.080 --> 19:35.400]  And awareness begins to get us back into sociology or history, right?
[19:35.460 --> 19:37.380]  Like those are sort of more about human questions.
[19:37.620 --> 19:39.080]  Back over there by the window, yes.
[19:41.320 --> 19:41.760]  Sorry?
[19:44.740 --> 19:45.960]  How do you know?
[19:46.660 --> 19:48.460]  If you use Skype, your chats are encrypted?
[19:50.040 --> 19:50.680]  Somebody's trolling.
[19:50.840 --> 19:51.640]  Okay, pink hair, yes.
[19:56.090 --> 19:57.110]  Monitor, mm-hmm, yes.
[19:57.250 --> 19:58.990]  We're going to get to that one shortly, yes.
[19:59.710 --> 20:00.030]  Yes.
[20:00.390 --> 20:01.850]  This is one of the things I found out at Second Life.
[20:01.930 --> 20:05.590]  The number of things that are counted about your behavior online are really kind of fascinating and interesting.
[20:05.770 --> 20:06.450]  We'll say a bit more.
[20:06.570 --> 20:11.130]  So I do still like the idea of like the idea of access is good, but then we also have the question of awareness.
[20:11.230 --> 20:12.490]  One more back by the cameras there, yeah.
[20:13.590 --> 20:23.590]  If you have a webmail service, you can offer different description options to ways to be selected to sets of people and see which ones get more updates.
[20:23.690 --> 20:25.310]  Okay, so that's more of an A-B study, right?
[20:25.490 --> 20:38.010]  So if you have an email service of some sort, you basically put the option to say, pop a little popup that says, hey, you can encrypt your email to some people and not to others.
[20:38.010 --> 20:44.750]  Or just maybe change the color of blue, like the way Google's changed, like they tried out 15 colors of blue and to figure out which one worked best.
[20:44.870 --> 20:46.330]  Like that's super quant, right?
[20:46.410 --> 20:51.430]  We're really getting into counts of who has picked things up.
[20:52.050 --> 20:53.590]  Then how do we know why?
[20:57.710 --> 20:58.610]  Multiple choice surveys.
[20:58.610 --> 20:58.810]  It's right.
[20:58.950 --> 21:00.610]  They're more quantitative than...
[21:00.610 --> 21:05.150]  So you're already giving people an answer and then you can just simply count the number of responses a particular way.
[21:05.210 --> 21:09.570]  So that begins to sort of simplify the number of variables, especially when you get something written out.
[21:10.830 --> 21:20.690]  Once you've gone through enough research training that you've had to like codify people's responses when they have open-ended responses, the more you're like, oh, God, I'm never doing this again, right?
[21:20.690 --> 21:22.470]  Because that takes a lot of work.
[21:22.590 --> 21:32.030]  And the other thing that it takes is a lot of going in and you have to define with your other researchers, is this an example of what we're trying to count here?
[21:32.150 --> 21:36.710]  So somebody says, yes, I use Skype, for example.
[21:36.890 --> 21:38.710]  Do we count that as using encryption or not?
[21:39.110 --> 21:41.870]  Because are they actually aware that there's encryption in it?
[21:41.930 --> 21:42.730]  Are they not aware?
[21:43.030 --> 21:43.370]  Right?
[21:43.490 --> 21:49.070]  So you then begin among your researchers to define what counts as evidence and what doesn't.
[21:49.070 --> 21:56.690]  So we'll talk a little bit more about having the researcher define that versus having the people you're studying define that in a moment.
[21:56.870 --> 22:02.410]  Because that's kind of, those are slightly two different things and they actually end up being in two different fields of social study.
[22:03.510 --> 22:03.870]  Right.
[22:04.010 --> 22:13.770]  So we, so I think one of the things we sort of haven't said yet is you can sort of, we said with the A-B test, you can see whether they picked it up if they, if they got this kind of pop-up versus that kind of pop-up, right?
[22:14.910 --> 22:16.890]  But that still doesn't quite tell us why.
[22:16.890 --> 22:20.310]  All we know is, I mean, it could be correlation, not causation, right?
[22:20.430 --> 22:29.750]  I mean, if it's, if it's sheerly between like, yes, you have a pop-up that says encrypt versus you had no pop-up, that doesn't necessarily, I mean, that tells you that yes, if you have a pop-up, they'll find it.
[22:29.810 --> 22:34.670]  But if you have two similar pop-ups, you begin to sort of not have the answer of why, like, why is this happening?
[22:34.710 --> 22:35.830]  I'm seeing way in the back of hand.
[22:35.990 --> 22:36.130]  Yes.
[22:45.760 --> 22:46.400]  Random assignment.
[22:52.120 --> 22:54.800]  Right, so yes, you can begin to control a lot of variables.
[22:54.960 --> 23:00.040]  That's one of the sort of nice things about the net is that you can sort of do that.
[23:00.280 --> 23:05.700]  I'll sort of, the problem is I have sort of multiple tracks I could go on on this particular line of thought.
[23:05.780 --> 23:08.340]  But we'll get to more talking about A-B testing in a second.
[23:08.960 --> 23:16.040]  So the point is generally that like, once again, especially if you're sort of controlling the environment, you're still not totally sure why people are choosing one thing or why they're not.
[23:16.880 --> 23:19.060]  You really have to begin to sort of talk to them about that.
[23:20.760 --> 23:30.340]  We could, for example, let's see, if we are, if we were going to try to find out why, what could we do to find out why?
[23:30.480 --> 23:31.380]  What other things could we do?
[23:31.500 --> 23:32.020]  Yes, right there.
[23:39.800 --> 23:42.700]  So sit with somebody in a room, probably in a lab, right?
[23:42.780 --> 23:46.600]  Or maybe even at their house, whatever, and observe them as they, as you said it, and you'd try to set it up.
[23:47.100 --> 23:49.900]  How do we know then if they are going to encrypt again?
[23:50.840 --> 23:50.920]  Right?
[23:51.100 --> 23:53.500]  So there are certain things you can know within a lab.
[23:53.600 --> 24:00.880]  You can certainly know whether your interface is working, but it's really hard to predict behavior going forward because the lab is sort of an unusual situation.
[24:01.020 --> 24:01.700]  Way back in the back again.
[24:01.780 --> 24:01.920]  Yes.
[24:08.580 --> 24:08.940]  Ah.
[24:09.940 --> 24:10.300]  Yes.
[24:15.420 --> 24:15.780]  Exactly.
[24:15.940 --> 24:17.120]  That's a very good point, right?
[24:17.340 --> 24:18.840]  So you can create a need to encrypt.
[24:18.900 --> 24:21.300]  And especially if you're in the lab, there are dynamics in the lab, right?
[24:21.360 --> 24:28.300]  Like a lab experiment on a human being is going to be in the context in the frame, as Irving Goffman says, of this is an experiment.
[24:28.480 --> 24:30.060]  The researcher wants something of me.
[24:30.180 --> 24:31.680]  What does the researcher want, right?
[24:31.800 --> 24:37.360]  So that creates an entirely different setting that the person might be reacting to, not in a way that's at all natural, right?
[24:37.780 --> 24:39.860]  But need is also an interesting word.
[24:40.000 --> 24:44.460]  Like if we go out and observe people, we can begin to find out like what do they say their need is, right?
[24:44.460 --> 24:50.140]  One of the other questions is like, what would the marketing team do if you ask the marketing team about encryption?
[24:52.020 --> 24:54.280]  What is the tool of the marketing team, for example?
[24:55.440 --> 24:55.880]  Hmm?
[24:57.280 --> 24:58.320]  They hate encryption.
[24:58.800 --> 25:00.100]  They're like, I can't sell this.
[25:00.180 --> 25:01.120]  It's not social media.
[25:01.360 --> 25:01.720]  Right?
[25:01.880 --> 25:06.240]  So if you think of like what marketing teams do, what is a tool that they use research-wise?
[25:08.060 --> 25:08.860]  Focus groups.
[25:09.060 --> 25:09.180]  Yes.
[25:09.400 --> 25:09.880]  Focus groups.
[25:09.980 --> 25:10.240]  Exactly.
[25:10.760 --> 25:16.580]  What is the problem with getting a lot of people in the room in a focus group and saying, so what do you think about encryption?
[25:17.420 --> 25:18.080]  What's wrong with that?
[25:18.160 --> 25:18.300]  Yes.
[25:19.000 --> 25:22.940]  Most people who are using crypto have no idea they're using it.
[25:23.220 --> 25:23.540]  Uh-huh.
[25:24.160 --> 25:24.600]  Oh, okay.
[25:24.700 --> 25:26.280]  My mother uses iMessage.
[25:26.320 --> 25:26.700]  Uh-huh.
[25:26.760 --> 25:35.200]  Because by her fault, it, you know, her iPhone upgrades text messages to other iPhones to, okay, it's not perfect, but it isn't.
[25:35.540 --> 25:35.560]  Right.
[25:36.200 --> 25:39.550]  Mom had no idea.
[25:39.750 --> 25:39.990]  Okay.
[25:40.130 --> 25:42.050]  And then what I did point it out, she had no idea.
[25:42.050 --> 25:42.490]  Okay.
[25:42.670 --> 25:53.210]  So that's, I mean, so a lot of times it's behind the scenes, but if we were talking to people, you know, about asking, if we're sort of more asking them, the major flaw there is, so part of it is they might not be aware, right?
[25:53.390 --> 25:55.150]  What else might they not be aware of?
[25:55.310 --> 25:55.390]  Yeah?
[25:55.550 --> 25:58.150]  To put them in a group, you don't get a consensus.
[25:58.870 --> 25:59.130]  Yes.
[25:59.990 --> 26:00.290]  Yeah.
[26:00.370 --> 26:01.550]  So that's the major problem with focus groups.
[26:01.690 --> 26:07.430]  Don't let your marketing team tell you to do focus groups because not only are they there reacting to the researcher, they're also there reacting to each other.
[26:07.590 --> 26:11.230]  So you're not necessarily, focus groups are really not the greatest tool for pretty much anything.
[26:12.430 --> 26:12.710]  Right.
[26:14.010 --> 26:16.850]  So, yeah, because self-report data is also unreliable.
[26:16.970 --> 26:20.690]  Like when people tell you I am going to do this or I did do this, frequently they're not.
[26:20.810 --> 26:27.430]  I mean, so a lot of the times people misestimate what they do most often, particularly if you ask them how often do you do something, that's not something they do all that commonly.
[26:29.030 --> 26:32.130]  So, just sort of going on with more and more ways of thinking about things.
[26:32.270 --> 26:35.750]  This is something you can observe about my secure tool use.
[26:35.830 --> 26:40.250]  This is my contact list in Jitsi, which is a tool that supports encrypted voice calls.
[26:41.090 --> 26:44.990]  What kind of research questions could we generate about encrypted tool use based on what we see here?
[26:47.210 --> 26:48.070]  Phrase me a question.
[26:49.070 --> 26:50.830]  I'm asking you guys to do an awful lot of work today.
[26:51.050 --> 26:51.110]  Yes?
[26:52.110 --> 26:52.430]  Encryption.
[26:54.680 --> 26:55.500]  Change the shade of...
[26:55.500 --> 26:58.060]  Yeah, but why would we think the shade of blue would change their encryption?
[26:58.240 --> 26:59.280]  We've been talking about color.
[26:59.400 --> 27:01.680]  We've been talking about how to indicate to people whether their message is secure.
[27:01.780 --> 27:03.060]  It's like, which color should we use?
[27:03.160 --> 27:05.800]  But blue is in max and then like, is red bad?
[27:05.940 --> 27:07.120]  But then there's red-green colorblindness.
[27:07.180 --> 27:11.340]  I think honestly the color fight is like my least favorite UX thing because I'm like, there are no true answers.
[27:12.160 --> 27:19.080]  Some countries, some languages only have words for like dark and light or like they don't have a distinction between blue and green.
[27:19.260 --> 27:20.960]  So like you don't even want to mess with that, I don't think.
[27:20.980 --> 27:21.560]  I think that's too much.
[27:21.720 --> 27:24.920]  What else could we ask about this particular thing you see right here?
[27:24.980 --> 27:25.160]  Yes?
[27:25.460 --> 27:27.040]  How do you know the people who are?
[27:28.140 --> 27:30.860]  You could ask the person who's encrypting, how do you know who they are?
[27:31.140 --> 27:32.740]  That is an interesting question, yeah, actually.
[27:32.860 --> 27:34.540]  So that's one thing we could observe about that there.
[27:34.620 --> 27:35.540]  Anything else that we could ask?
[27:36.180 --> 27:36.500]  Yes?
[27:39.490 --> 27:41.170]  How do I know it's doing what it says it's doing?
[27:41.170 --> 27:46.330]  That is actually something maybe more that a user would ask of the researcher, but that's actually a really good question.
[27:46.470 --> 27:48.210]  Like how does... you could ask the user, how do you know?
[27:48.310 --> 27:49.470]  Like how do you know it's encrypting, right?
[27:49.730 --> 27:53.310]  I think my favorite question about like this is the entirety of my SIP contact list, right?
[27:53.610 --> 28:02.370]  Because one of the major problems we're facing in encryption is if you need to have a trusted network and you need to create it out of thin air, do you suck down your Google contacts?
[28:03.070 --> 28:06.050]  Then you have all your Google contacts that might not be safe.
[28:06.210 --> 28:08.450]  So like this is a major challenge.
[28:09.030 --> 28:10.610]  But once again, if you sort of...
[28:10.610 --> 28:11.650]  So we could do an example.
[28:11.750 --> 28:19.150]  We could do a test where we sort of go, okay, have the server for Jitsi ping the software and go, what is the end of the contact list?
[28:19.210 --> 28:19.870]  How many people are on there?
[28:19.930 --> 28:21.170]  We don't even need to know who they are, right?
[28:21.910 --> 28:26.790]  But it would be good to have some sort of, you know, set...
[28:26.790 --> 28:31.490]  We could begin to then make correlations between like, if somebody is not using the tool that much, is it related to this?
[28:31.570 --> 28:35.930]  Or if they simply cease using the tool for long periods of time, is it because they just don't have that many people on their contact list?
[28:36.250 --> 28:37.670]  Sorry, I'm realizing I'm sort of falling behind.
[28:38.390 --> 28:48.010]  I think we've to some extent determined, and I'm sort of making a leap here because I think we haven't quite gotten quite got there, but we decided that A-B testing would be pretty cool to do, right?
[28:48.150 --> 28:54.830]  Like A-B testing is a way that you can actually get a sense of people's behavior without bringing them in.
[28:54.930 --> 29:03.550]  So A-B testing, once again, is popping up something on their machine or having your software send you information about events like, did they log in, did they log out?
[29:04.030 --> 29:05.210]  How long have they been on?
[29:05.350 --> 29:07.030]  You know, other pieces of information like that.
[29:07.630 --> 29:12.190]  And the Internet makes this kind of observing of actual human behavior really easy.
[29:12.210 --> 29:13.570]  So we don't have to do a focus group.
[29:13.630 --> 29:15.530]  We don't have to get them to give us self-report data.
[29:15.650 --> 29:17.150]  We actually find out what they do.
[29:17.430 --> 29:19.230]  All the major companies do it all the time.
[29:19.350 --> 29:20.370]  Google does it, Facebook do it.
[29:20.430 --> 29:23.270]  They have very, very sexy data if you're a social researcher, right?
[29:23.350 --> 29:24.650]  Yes, sexy data, yes.
[29:24.650 --> 29:30.490]  So, but of course, we recently found out where that ends up, right?
[29:30.770 --> 29:34.650]  So, I don't know, how many people heard about the Facebook emotional manipulation study?
[29:34.810 --> 29:35.990]  Yeah, most of the people in the room, right?
[29:36.030 --> 29:36.690]  Because that's why you're here.
[29:37.630 --> 29:44.150]  Facebook did emotional manipulation studies on users by selecting what messages they see on the wall by emotional tone.
[29:44.150 --> 29:51.750]  So they would only show you the friends who are saying happy things, or they would only show you the friends that were saying sad things, and then they looked at how that impacted people's behavior.
[29:53.090 --> 29:55.030]  Needless to say, this kind of defeats the purpose.
[29:55.170 --> 30:00.230]  If you're looking at security tools, it kind of defeats the purpose of privacy protection, and also of ethics, right?
[30:00.450 --> 30:07.710]  Like, there's a lot of discussion in certain groups that I'm in of, you know, should...
[30:07.710 --> 30:14.010]  Because in academia, if you're in the ivory tower, if you're in a university, you have to go to a group called an institutional review board.
[30:14.210 --> 30:16.870]  And they tell you, yes, you're doing things that are safe.
[30:16.950 --> 30:17.650]  No, you are not.
[30:17.810 --> 30:22.870]  So they're the ones who say, no, you may not administer electric shocks to pregnant women to find out if, you know, whatever happens.
[30:23.650 --> 30:24.930]  Yeah, and they're very, very strict.
[30:25.090 --> 30:26.030]  Like, IRBs are quite strict.
[30:26.130 --> 30:30.710]  And that's sort of one of the reasons I think Facebook gets to have sexier data, is there's nobody telling them no, right?
[30:31.570 --> 30:40.530]  But for us, for those of us who are in a community that cares a lot about protection of privacy, security, anonymity, these things are sort of difficult.
[30:40.670 --> 30:44.430]  And we're a little bit back to square one about how to do that research.
[30:44.850 --> 30:46.730]  That is kind of a whole other talk.
[30:46.730 --> 30:50.030]  I think it will be discussed in usability talks tomorrow.
[30:50.490 --> 30:53.210]  And we also have a usability meetup tomorrow that I want to make a plug for.
[30:53.350 --> 30:53.890]  Come and join us.
[30:53.970 --> 30:55.850]  We're going to talk about usability and secure tools.
[30:57.970 --> 31:01.270]  Yeah, so we have a lot of reckoning to do about AB tests.
[31:01.470 --> 31:03.010]  Ooh, hey, I have a double with that slide.
[31:03.150 --> 31:03.790]  I don't know why that happened.
[31:03.950 --> 31:04.190]  Ha, ha, ha.
[31:05.630 --> 31:05.950]  Right.
[31:07.550 --> 31:09.630]  Okay, we've talked over all that part already.
[31:09.930 --> 31:10.970]  I moved my slides around.
[31:11.110 --> 31:11.510]  My apologies.
[31:12.570 --> 31:20.410]  Anyway, by the way, any time you see a study reported in the news about user behavior, about human behavior, you should be asking these questions about the research.
[31:20.570 --> 31:22.610]  You should be asking, how did they do this?
[31:22.630 --> 31:23.650]  How did they ask the question?
[31:23.950 --> 31:25.310]  How did they get their data?
[31:25.590 --> 31:26.790]  Do I believe it's valid?
[31:27.170 --> 31:27.450]  Right?
[31:27.530 --> 31:31.590]  Do I believe that this was a valid way to get the right answer about this kind of thing?
[31:32.770 --> 31:42.750]  What I'm trying to get at here is that simply doing, as you probably know by now, I'm trying to get at the point that quantitative mathematical science is simply not the right tool for the job.
[31:43.010 --> 31:43.030]  Right?
[31:43.150 --> 31:46.710]  It's not the right tool for asking questions about why human beings do things.
[31:46.770 --> 31:51.790]  It's a great tool, but it doesn't usually tell you what people's logic and reasoning is themselves.
[31:53.930 --> 31:56.970]  And this is where I mean your belief in science is wrong.
[31:56.970 --> 32:09.450]  If you find yourself believing that doing a more scientific study is better, if a more statistical study is better than an anecdotal study or an observational study, you're doing science wrong.
[32:09.570 --> 32:09.650]  Right?
[32:09.830 --> 32:12.790]  Because just doing math to it doesn't make it a better science.
[32:12.950 --> 32:13.010]  Right?
[32:13.850 --> 32:15.910]  Science doesn't ask that you believe in it.
[32:15.990 --> 32:17.470]  Science asks that you interrogate.
[32:17.830 --> 32:18.110]  Right?
[32:18.230 --> 32:19.010]  That is the idea.
[32:19.170 --> 32:19.270]  Right?
[32:19.330 --> 32:25.230]  So you want to make sure that you are asking questions the right way and then asking questions of a report that you see.
[32:26.030 --> 32:28.410]  And I've got a very famous physicist backing me up on this one.
[32:29.670 --> 32:34.270]  So, has anybody read Richard Feynman's Cargill Cult Science paper?
[32:35.010 --> 32:35.710]  Essay that he did.
[32:35.850 --> 32:38.570]  It was a convocation to Caltech students.
[32:39.370 --> 32:41.770]  He called out things that pretend to be science, but aren't.
[32:41.850 --> 32:44.430]  And he actually cites my field educational research as an example.
[32:44.550 --> 32:46.670]  And I said, ow, my very expensive degree.
[32:48.110 --> 32:49.870]  But I actually, I agree with Feynman.
[32:49.930 --> 32:50.310]  I really do.
[32:50.390 --> 32:52.190]  There's a lot of bad educational research out there.
[32:52.190 --> 32:55.170]  And a lot of really terrible research about people and cultures.
[32:55.410 --> 33:00.370]  One of my least favorites is the neurocognition studies on ethics.
[33:00.770 --> 33:03.590]  We can tell you what an ethical part of brain is.
[33:03.790 --> 33:06.310]  You know, it's like caveman questions, I think, a lot of the time.
[33:07.930 --> 33:11.970]  I think frequently the more a social research study pretends to be hard science, the worse it is.
[33:12.030 --> 33:14.050]  And I think that was basically what Feynman was calling out, right?
[33:14.110 --> 33:16.810]  If you're trying, if you're doing it the wrong way, then yes, it's bad science.
[33:17.990 --> 33:19.610]  Because they're not the right methods for this thing.
[33:19.610 --> 33:25.350]  So, I like to argue that actually being in the field and observing people is more empirical, right?
[33:25.410 --> 33:30.030]  You're actually observing a thing that happens when you're studying culture than bringing people into a lab, right?
[33:32.170 --> 33:40.190]  For my, in full disclosure, my opinion is that for many questions, the logic of anthropology is generally more rigorous than sociology or psychology.
[33:41.250 --> 33:49.950]  Anthropology, this is Franz Boas, or Franz Boas, terrible visual pun, the founder of anthropology, as we know it.
[33:50.890 --> 34:03.110]  Anthropology holds that the observation of multiple minds in communication, behavior, conversing with multiple minds, and participation in people's daily life is the strongest means of understanding people's rationale for what they do.
[34:05.330 --> 34:09.730]  Laboratory tests promote, purport to eliminate variables from social interaction.
[34:10.590 --> 34:12.670]  But think about it, that's actually kind of counterproductive, right?
[34:12.770 --> 34:14.750]  You don't necessarily want to eliminate variables.
[34:14.870 --> 34:16.750]  You want to know what is actually happening, right?
[34:16.830 --> 34:18.610]  You're looking for social variables.
[34:18.610 --> 34:24.010]  And when you have a lab study, once again, you introduce the variable of, like, the researcher wants me to do this, right?
[34:28.110 --> 34:30.270]  And participation is a big part of it, too.
[34:30.390 --> 34:35.910]  A lot of the time, what you're doing as an anthropologist is going out and observing, but also doing what people do.
[34:36.110 --> 34:47.170]  You're being a daughter in an Eskimo community and finding out what is right and wrong to do as a daughter, so you begin to learn as a person reacting to, you know, the people around you what is correct and what is incorrect.
[34:47.750 --> 34:53.330]  If your question is why are people asking like this, you need to observe what people also tell each other, right?
[34:53.330 --> 34:55.010]  So about what is correct and not correct.
[34:55.710 --> 34:58.510]  You need to participate in and break how they're acting with each other.
[34:58.870 --> 35:01.110]  And you need to ask people why they're behaving like this.
[35:02.550 --> 35:07.370]  You need to know the reasoning that people are following when they behave a certain way and when they're reinforcing each other's choices.
[35:07.570 --> 35:10.330]  And remember, oh, we haven't quite got to that part.
[35:10.470 --> 35:15.250]  So anthropology and sociology are the studies of people interacting with each other.
[35:16.370 --> 35:28.770]  Psychology tends to be the study of one head, like what is going on in a single person's head, which may be why I think earlier I had a little box from Randall Munro saying, social psych is worse than either social or psychology, because I've always sort of felt like,
[35:28.790 --> 35:29.930]  which method are you using?
[35:30.030 --> 35:34.790]  Which question are you actually asking if you're thinking at the exact same time about one head and multiple heads?
[35:34.930 --> 35:36.110]  So anyway, that's a thing to work on.
[35:37.330 --> 35:47.250]  The idea of doing something wrong is actually a method of a sociology and anthropology subfield called ethno-methodology, which is a very long name for a field.
[35:47.250 --> 35:51.150]  The founder also has the very long name of Harold Garfinkel.
[35:51.750 --> 35:55.110]  And his writing is among the most tortuous and hard to read, so I'll sum it up for you.
[35:56.150 --> 35:57.650]  It's really like unreadable.
[35:58.650 --> 36:06.370]  But if you... his basic idea is if you break a social convention, that is the best way to learn what the social convention is.
[36:06.550 --> 36:13.570]  Because if you say... say you go into a bank and you cut in line, everybody will correct you and say, hey, no, the line is over there, the end of the line is over there, right?
[36:13.750 --> 36:16.690]  So actually getting in and disrupting what's going on.
[36:16.770 --> 36:19.570]  And actually this gets done in user tests as well in computer science.
[36:19.630 --> 36:25.050]  If you suddenly make a window behave in a way that people don't expect, you begin to learn a little bit more about what they did expect, right?
[36:25.150 --> 36:26.250]  So that sort of tells you those things.
[36:28.810 --> 36:33.330]  So anthropology and ethno-methodology, these observational fields, are empirical, right?
[36:33.430 --> 36:36.510]  They are generalizing about things that happen reliably.
[36:36.530 --> 36:37.750]  They happen over and over again.
[36:37.850 --> 36:41.010]  You don't just as an anthropologist say, oh, I saw this person do this once, right?
[36:41.010 --> 36:43.610]  You say multiple times this happened.
[36:44.150 --> 36:46.970]  I asked somebody about it and they said, yes, that's how it works.
[36:47.170 --> 36:51.590]  And then I interrupted that process and somebody else interrupted that process and they were corrected, they told them the right way.
[36:51.930 --> 36:56.410]  These are not fields that just employ statistics or laboratory regulation to achieve validity, right?
[36:57.030 --> 37:10.670]  The difference between the social sciences and the so-called hard sciences is the difference between a logic that operates on what is observably factual in the world, right?
[37:12.530 --> 37:13.170]  Anthropology... sorry.
[37:13.730 --> 37:16.290]  That was the logic in the... logic in the hard sciences.
[37:16.470 --> 37:17.750]  This is where I'm beginning to melt down a little bit.
[37:18.250 --> 37:19.530]  The logic in the hard sciences, right?
[37:19.650 --> 37:23.910]  There is a logic that operates on what is observably factual in the world.
[37:24.010 --> 37:30.010]  When you go and you do a chemistry experiment, you expect it to operate by the same logic next time, or biology experiment, or mathematical proof.
[37:30.890 --> 37:32.770]  They all do the same things the same way.
[37:34.590 --> 37:44.810]  The soft sciences, anthro, soci, and psych, by contrast, are the observable fact of a logic, of people's logic, operating on the world, right?
[37:44.990 --> 37:50.710]  And the logic itself is the object of observable, of qualitative study, right?
[37:50.750 --> 37:56.930]  It's in people's heads and actions, and it's just not effective to try to study it as if it was a quark or an equation or an endothermic reaction.
[37:58.050 --> 38:06.690]  The people comp sci encountered in Fulani do repeatedly apply logic, both in deciding where not to step, and in deciding what to do when the computer isn't reacting the way they expect.
[38:07.310 --> 38:11.310]  And asking the teacher was not a bad strategy there, he just didn't like that strategy.
[38:11.470 --> 38:13.030]  I was like, why is that not okay with you?
[38:13.070 --> 38:13.650]  They're your students.
[38:14.570 --> 38:16.850]  And Greenwald was applying the logic of a busy journalist.
[38:16.950 --> 38:23.130]  He didn't know how big Snowden's story was, and encryption seemed like it was going to take a long time, so he just decided not to take that time, right?
[38:23.170 --> 38:24.470]  It was a relatively logical thing to do.
[38:24.470 --> 38:28.190]  It's not your logic, right, because you know more about encryption and what it does.
[38:28.410 --> 38:36.150]  It's not a scientific logic, but you'd better believe that computer users who don't share your culture are going to apply it again and again, probably the same way every time.
[38:36.330 --> 38:41.750]  And as such, it becomes a fact that you must reckon with as a software developer and as a scientist.
[38:43.090 --> 38:54.250]  Unfortunately, I see coders in the open-source and security tool communities acting as if they are unaware that their so-called logical assumptions, their desire to do math and science to everything, carries their own cultural logic with it.
[38:54.370 --> 38:55.150]  It's not pure.
[38:55.710 --> 38:56.650]  It's not neutral.
[38:57.670 --> 38:59.370]  And software carries values with it, too.
[38:59.670 --> 39:04.330]  You may want to think that Linux is objectively the right way to run a computer, and Richard Stallman certainly does.
[39:04.710 --> 39:05.690]  Actually, no, not Linux.
[39:05.850 --> 39:06.370]  Sorry, GNU.
[39:07.650 --> 39:09.110]  He called me out on that already.
[39:09.210 --> 39:10.170]  We'll get to that in a moment.
[39:10.870 --> 39:16.370]  You may think that encryption rationally improves computer use, but neither of these beliefs are true in the abstract.
[39:16.930 --> 39:19.190]  There is no abstract in social behavior, generally.
[39:19.710 --> 39:24.770]  Being unaware of your own values and habits hurts the spread of the good work that we all want to do.
[39:26.830 --> 39:27.570]  Sorry, yes?
[39:27.950 --> 39:30.030]  Aren't morals like abstract?
[39:31.390 --> 39:32.390]  We can get to that.
[39:32.470 --> 39:36.830]  I would say morals are more concrete in the world, but that is a question for after.
[39:37.550 --> 39:38.030]  I'm sorry.
[39:38.150 --> 39:41.270]  I'm now in the, like, I need to finish this, and then we can get to discussion afterwards part of the talk.
[39:42.150 --> 39:51.770]  Evan Moglen, the writer of the GNU Public License, gave an incredibly appealing talk at the previous HOPE, in which he used vampires as a metaphor for Apple products and Steve Jobs.
[39:53.010 --> 39:53.930]  It was great.
[39:54.070 --> 39:54.450]  You should look it up.
[39:54.490 --> 39:57.210]  It was really just one of the most beautiful speeches I've ever heard.
[39:57.470 --> 40:05.850]  He said, their goal is to seduce through their incredible beauty and then to make us dependent, to make us just like them, to sap our precious bodily fluids.
[40:05.850 --> 40:06.710]  Right?
[40:07.090 --> 40:11.410]  I thought Moglen's ideas were beautifully phrased, and I agreed with him for most of it.
[40:12.290 --> 40:20.870]  But then I started working more closely with the open-source community, and I was struggling daily to get a Linux computer to work without being completely dependent on somebody else to get anything done.
[40:21.290 --> 40:26.250]  And then I heard a lot of people in that community, in the open-source community, repeating Moglen's themes.
[40:26.510 --> 40:32.550]  When I asked for a Mac, I was dismissed for wanting something shiny or something pretty.
[40:33.150 --> 40:35.230]  It was basically the same words that Moglen used.
[40:35.330 --> 40:36.250]  He also said jewelry.
[40:36.690 --> 40:42.130]  So then I began to think maybe what I was hearing was Macs are for girls and Linux is not.
[40:42.310 --> 40:45.750]  And because these computers don't work for you, you don't belong here.
[40:47.290 --> 40:48.930]  But then I thought about it a little bit more.
[40:49.030 --> 40:50.030]  I don't think it's just gender.
[40:50.130 --> 40:52.030]  I don't think that is actually what I was hearing.
[40:52.030 --> 40:57.950]  I think the open-source dogma that downplays the value of aesthetics is more like a fundamentalist religion.
[40:58.190 --> 40:59.670]  Moglen wants you to be Amish.
[41:05.780 --> 41:07.660]  That's Jason Scott's brand on the cow.
[41:07.800 --> 41:09.000]  I was told I had to put that in.
[41:09.500 --> 41:11.120]  He wants you to not have zippers.
[41:11.260 --> 41:15.180]  He wants you to wear modest necklines and sleeves that are down to here, right?
[41:15.480 --> 41:22.660]  Because those aesthetics... because the aesthetics of the Macintosh are the false seductions of a modern world of computing that does not care about your immortal soul.
[41:24.020 --> 41:24.500]  Right?
[41:25.340 --> 41:30.400]  Having a rigid cultural code is really good for creating an insular little religious community, right?
[41:30.480 --> 41:32.560]  It's not great for spreading your views to the world.
[41:32.860 --> 41:37.020]  One of the most successful things about Christianity is that it is flexible, right?
[41:37.360 --> 41:38.880]  Amish sects are not that flexible.
[41:39.220 --> 41:43.720]  The majority of Christianity allows itself to go out in the world and be adapted into all kinds of places, right?
[41:45.300 --> 41:51.360]  The code that Stallman and Moglen insist on is cultural or even... it's moral.
[41:51.360 --> 41:52.560]  It is very deeply moral.
[41:52.660 --> 41:53.480]  There are morals that go with it.
[41:53.600 --> 41:55.240]  It's not scientific or objectively true.
[41:55.940 --> 42:03.260]  And I wouldn't mind so much that they're policing aesthetics except that oftentimes they don't make the distinction between what is pretty and shiny and what is usable.
[42:03.580 --> 42:05.720]  And most users can put up with a lack of pretty.
[42:05.720 --> 42:09.880]  They cannot put up with a lack of usable and doing everything from the command line.
[42:11.760 --> 42:12.280]  All right?
[42:13.660 --> 42:16.260]  So this is why I'm saying solve the hard problem, right?
[42:16.400 --> 42:19.520]  What if the hard problem isn't cool math problems or physics problems?
[42:19.880 --> 42:24.000]  What if it's really world peace or world hunger or resource distribution generally?
[42:25.840 --> 42:32.600]  CompSci, what if the hard problem isn't making a better encryption algorithm but making encryption tools easier to use and understand?
[42:32.600 --> 42:36.020]  This is from, I think, either Enigmail or GPG.
[42:36.640 --> 42:37.960]  That's my fingerprint, look me up.
[42:39.140 --> 42:42.060]  So much harder than saying this is my at on Twitter.
[42:42.900 --> 42:45.940]  What if the hard problem isn't making a Linux distro that works for you?
[42:46.220 --> 42:50.700]  What if it's making a Linux distro that never requires a computer to work from the command line ever?
[42:51.740 --> 42:57.580]  What if the hard problem is learning to talk to people you decided a long time ago weren't worth your time because they weren't smart?
[43:00.640 --> 43:01.720]  Okay, shut up, you're saying.
[43:01.860 --> 43:04.260]  I may be not so good with the people part sometimes, right?
[43:04.380 --> 43:04.960]  What do I do?
[43:05.140 --> 43:06.180]  Okay, I've got some suggestions.
[43:06.380 --> 43:06.540]  There we go.
[43:07.720 --> 43:11.740]  First of all, are you deciding whether you should devote your time to a new software project?
[43:12.320 --> 43:15.600]  First of all, don't make your own new security tool.
[43:15.600 --> 43:17.380]  It will not go as well as you think, right?
[43:19.220 --> 43:20.440]  Go to an existing project.
[43:20.560 --> 43:24.220]  There are so many crypto tools out there that really need your help.
[43:25.140 --> 43:26.460]  MailVelope is one of my favorites.
[43:26.680 --> 43:31.080]  It's this part of the ecosystem where it works with browser-based email and we'll encrypt that.
[43:31.660 --> 43:33.660]  It's just easier for a lot of people to use.
[43:33.820 --> 43:39.320]  And so in a harm reduction model saying we're improving things slightly, MailVelope is a good thing to use.
[43:39.620 --> 43:40.520]  Go help MailVelope.
[43:41.000 --> 43:43.800]  Enigmail, KeePass, and Jitsi could also use a bit of developer love.
[43:44.860 --> 43:49.340]  Contribute code that specifically makes things easier for non-technical people to use, right?
[43:49.340 --> 43:51.180]  This is the commotion project.
[43:51.420 --> 43:58.860]  Their GitHub repo has... or the GitHub site has usability tags on it so you know actually which things are bugs that are usability so you can actually go and fix them.
[43:59.060 --> 44:03.700]  If you have a project, add usability tags to saying like this is a usability thing and then go and fix those.
[44:04.120 --> 44:05.240]  Have people help you fix those.
[44:06.360 --> 44:09.420]  Include user researchers and designers on your team if you're developing software.
[44:09.980 --> 44:12.820]  Work with them to... when you make a new app or site, right?
[44:13.780 --> 44:15.980]  Maybe even listen to your marketing team sometimes.
[44:16.740 --> 44:20.140]  But just don't tell them... don't let them tell you to use focus groups because focus groups are bogus, right?
[44:20.520 --> 44:24.780]  You can... I mean now that you've been through this talk, you can be like let's talk about good methods for finding things about people, right?
[44:24.880 --> 44:26.000]  So you can sort of work together on that.
[44:26.580 --> 44:30.160]  Learn more about qualitative methods and which ones are good for which questions.
[44:32.120 --> 44:34.780]  I like Jakob Nielsen's alert box.
[44:34.920 --> 44:35.800]  It's not all qualitative.
[44:36.000 --> 44:36.900]  There's a lot of good stuff in there.
[44:37.760 --> 44:42.700]  Look a little bit more into ethnography and why it's valid, why it does what it does, and into ethnomethodology.
[44:43.340 --> 44:47.280]  On a really basic human level, get better at asking questions of other people, right?
[44:48.780 --> 44:51.120]  Noting again, do you encrypt your messages?
[44:51.480 --> 44:54.260]  What is the implicit assumption in the question, do you encrypt your messages?
[44:55.200 --> 44:56.300]  They know what encryption is.
[44:56.380 --> 44:56.680]  What else?
[44:57.580 --> 44:59.080]  That you should, right?
[44:59.280 --> 45:00.340]  That's a big one right there.
[45:00.540 --> 45:02.180]  So that's kind of a closed-ended question.
[45:02.240 --> 45:04.880]  This is one of the questions like how long have you been beating your wife, right?
[45:04.940 --> 45:06.080]  Like it's a very leading question.
[45:06.960 --> 45:10.740]  Ask and said something like what do you do to keep your communications safe and private?
[45:10.740 --> 45:12.680]  People already have practices that they're using.
[45:13.020 --> 45:20.280]  And if you can build on those and support those more and then go that extra mile and give them a little bit more support to do better crypto, that's good.
[45:21.740 --> 45:27.880]  Yeah, so asking non-leading open-ended questions is a good way to get people to tell you more about themselves generally also, just socially.
[45:28.560 --> 45:37.280]  Get to know a particular community of users like Muslims in the U.S. or LGBT activists in Russia, as seen here, and help ensure that their needs are heard in software development.
[45:37.900 --> 45:41.620]  LGBT communities have really, really interesting needs in terms of privacy and security.
[45:42.080 --> 45:43.840]  There's a real need to speak out, right?
[45:44.000 --> 45:49.060]  And I'm hearing from people who work with LGBT activists in Russia that they're like, they don't want to hide.
[45:49.200 --> 45:50.320]  They don't want to encrypt things.
[45:50.380 --> 45:51.080]  They don't want to get away.
[45:51.220 --> 45:53.380]  But at the same time, they do need to manage their physical safety.
[45:53.620 --> 45:56.040]  People need to not know where they are at particular points in time.
[45:56.320 --> 46:03.540]  So their speech needs to get out and then they need to protect their physical person, which is different than, for example, people under full surveillance in China.
[46:03.540 --> 46:05.020]  It's a very different kind of a model, right?
[46:05.900 --> 46:09.640]  Generally, talk to people who aren't like you and ask them what they need software to do.
[46:10.780 --> 46:11.220]  Thank you.
[46:11.320 --> 46:13.100]  And now we have five minutes for questions, I think, so.
[46:15.360 --> 46:15.980]  Good timing.
[46:16.200 --> 46:16.420]  All right.
[46:20.630 --> 46:22.050]  Amazingly, I don't feel sick anymore.
[46:22.510 --> 46:23.690]  Yes, standing up with the baby.
[46:26.370 --> 46:27.450]  Oh, I'm actually sorry.
[46:27.510 --> 46:29.170]  People are supposed to go to their microphones or no?
[46:29.450 --> 46:29.570]  No?
[46:29.710 --> 46:30.110]  Okay, just go ahead.
[46:30.330 --> 46:31.270]  There's a microphone over there.
[46:31.570 --> 46:32.590]  But go ahead for now.
[46:32.790 --> 46:43.530]  I'm wondering if it's worth taking your single axis and dividing it into two axes, one of which is the subject you're studying and the other is the methodology of study?
[46:43.770 --> 46:44.170]  Totes.
[46:44.430 --> 46:45.970]  You're an epistemology geek, I can tell.
[46:46.090 --> 46:46.250]  Hey, what?
[46:46.270 --> 46:46.990]  Repeat the question.
[46:46.990 --> 46:47.750]  Oh, yes, sorry.
[46:47.890 --> 46:54.830]  Is it worth splitting the chart into, it's actually not mine, it's Randall Munrose, into two axes, one about the methodology and one about the subject?
[46:54.930 --> 47:00.290]  I think that's a really good idea, particularly given that like more and more of what we do in life is multidisciplinary.
[47:00.990 --> 47:02.670]  And so, and methods really do travel.
[47:02.850 --> 47:05.970]  Like anthropologists will do, they'll do surveys.
[47:06.090 --> 47:07.990]  They'll do like quantitative counts of the community.
[47:08.250 --> 47:10.950]  And then sociologists will also do ethnography to observe things.
[47:11.210 --> 47:16.650]  So, yeah, I think just sort of identifying methods and then taking them across from place to place is a really good idea.
[47:16.870 --> 47:18.810]  So, yes, hand next to him there.
[47:19.370 --> 47:19.810]  Yeah?
[47:19.810 --> 47:20.950]  The guy in the hat.
[47:21.050 --> 47:21.170]  The guy in the hat.
[47:21.170 --> 47:21.590]  Oh, boy.
[47:22.870 --> 47:25.490]  My job is solving a really hard problem.
[47:25.650 --> 47:33.770]  I need to make all of the technology on earth accessible to computer disabilities, which is about 3% of the technology that's currently sold.
[47:42.130 --> 47:43.310]  It's super expensive.
[47:43.570 --> 47:46.970]  How do you teach 10-year-olds with no computer background experience?
[47:47.570 --> 47:52.470]  So, the hard engineering science problems are already solved.
[47:52.590 --> 47:56.370]  But the social problem is, how do we convince people?
[47:57.930 --> 47:59.030]  Everybody ought not.
[48:03.330 --> 48:03.710]  Right.
[48:04.330 --> 48:11.650]  Yeah, so the comment is the great point about disability usability and how hard it is to get those standards adopted everywhere.
[48:11.770 --> 48:12.710]  It's very, very important.
[48:12.810 --> 48:17.710]  It's actually important as much for activist communities as anybody else, because there's disabled folks in those communities as well.
[48:17.950 --> 48:20.270]  So, yeah, and that is a major challenge.
[48:20.490 --> 48:24.010]  We are having these discussions across encryption tech communities.
[48:25.010 --> 48:28.550]  And even when we have the discussions, you can't necessarily get everybody on the same page.
[48:28.550 --> 48:30.130]  And that is, yeah, it's a really hard problem.
[48:31.570 --> 48:42.290]  One of my major suggestions for anybody learning about social change is read Everett Rogers' Diffusion of Innovations, which is a massive compendium of all of the social change research that's ever been done.
[48:42.450 --> 48:45.750]  It's basically the book that Malcolm Gladwell stole from in order to write the tipping point.
[48:46.670 --> 48:54.790]  And so it's useful in thinking about how change happens, how people adopt things, who adopts first, what kind of messages need to happen.
[48:54.990 --> 48:56.790]  So that's a big recommendation for me.
[48:57.050 --> 48:57.250]  Yes?
[48:58.390 --> 49:07.910]  What are some projects that either have good outcomes or have good processes in terms of crypto, like current software projects that are delivering on some of these things?
[49:08.470 --> 49:11.150]  What current software projects have good usability specifically?
[49:11.450 --> 49:11.710]  Like which...
[49:11.710 --> 49:14.170]  Have a good process for these things.
[49:14.970 --> 49:20.030]  I think the teams that I'm thinking of that do really well, Commotion actually is doing a really good job with their documentation.
[49:20.850 --> 49:22.690]  They're on wireless and mesh networking software.
[49:24.230 --> 49:26.430]  The Whisper Systems team is doing a lot of work.
[49:26.510 --> 49:27.950]  Like when I said to them, what's your usability plan?
[49:28.030 --> 49:29.750]  They said, we're going to a local community center.
[49:29.890 --> 49:33.190]  I'm like, good job, because you're finding people who are really not like you.
[49:33.670 --> 49:37.610]  They're in the Bay Area, so that's really particularly hard, because it's just gentrifying so much.
[49:37.670 --> 49:42.990]  You can't find people who are not of the same educational background of you, maybe of a different culture than you and things like that.
[49:43.150 --> 49:44.310]  So their project is good.
[49:44.490 --> 49:45.430]  And I know I'm forgetting...
[49:45.430 --> 49:52.170]  Oh, Guardian Project has a large team that's doing a lot of integrating usability work from the beginning.
[49:54.770 --> 49:56.750]  And the other one I'm thinking of, Cryptocat.
[49:57.150 --> 49:57.510]  Cryptocat.
[49:57.650 --> 49:58.210]  There's Nadeem here.
[49:58.650 --> 49:59.550]  Go Cryptocat.
[49:59.790 --> 50:02.010]  Cryptocat's known for being pretty sleek and usable.
[50:02.010 --> 50:03.010]  So, yes?
[50:06.170 --> 50:13.690]  Wouldn't the usability problem be solved if there was market demand for these tools?
[50:13.850 --> 50:14.190]  Yes.
[50:14.470 --> 50:18.550]  Wouldn't the usability problem be solved if there was market demand for these tools?
[50:19.330 --> 50:20.410]  Yes, it would.
[50:20.690 --> 50:23.050]  And as a matter of fact, I think we're kind of seeing that.
[50:23.250 --> 50:29.130]  Google recently made some adjustment that made crypto easier for, I forget what it was,
[50:34.870 --> 50:39.750]  I'm speaking not from my job at all, speaking entirely personally, I think the best thing we could do would be protocols, right?
[50:39.870 --> 50:44.910]  Like get all of this into the protocol level and so people never, ever, ever, ever, ever have to listen to it or deal with it.
[50:45.250 --> 50:53.370]  But once again, that is a social problem because... and that's the same problem that we have with disability usability, is getting everybody on the same page for the protocols.
[50:53.550 --> 50:58.890]  And there will be corporate interests that will have particular takes on how it should go, and there will be you know, other smaller groups.
[51:00.550 --> 51:10.390]  It's hard to get adoption there, but that is certainly if you if you feel like there's something you should do and you're feeling about doing a software project, get involved maybe in helping people adopt new standards, because I think that might be a way to go to.
[51:10.590 --> 51:12.310]  So I'm told I'm at time.
[51:12.570 --> 51:13.590]  Thank you all very much.
[51:13.750 --> 51:16.430]  I can't believe there's this many people here on a Friday morning.
[51:17.070 --> 51:18.110]  Bless you all.
[51:18.610 --> 51:19.310]  Holy crap.
[51:19.450 --> 51:20.370]  Thanks for sticking around, guys.