Jofish Kaye on industry research, having an impact, and values-driven decision making

1

00:00:00,000 --> 00:00:01,000

CAL39_Jofish_Kaye



2

00:00:01,000 --> 00:00:02,000

Thu, Jan 26, 2023 • 1:07:40 mins



3

00:00:02,000 --> 00:00:03,000

http://www.changingacademiclife.com/blog/2019/2/2/jofish-kaye



4

00:00:03,000 --> 00:00:04,000

SUMMARY KEYWORDS



5

00:00:04,000 --> 00:00:05,000

mozil a, people, universities, work, industry, hci, impact, cal , chi, nokia, company, talk, important, world, absolutely, academia, bit, academic, papers



6

00:00:05,000 --> 00:00:06,000

SPEAKERS



7

00:00:06,000 --> 00:00:07,000

Geri, Jofish



8

00:00:07,000 --> 00:00:08,000

Geri:
00:04


9

00:00:08,000 --> 00:00:09,000

Welcome to changing academic life. I'm Geraldine Fitzpatrick. And this is a podcast series where academics and others share their stories, provide ideas and provoke discussions about what we can do individually and collectively to change academic life for the better.



10

00:00:09,000 --> 00:00:10,000

Geri:
00:30


11

00:00:10,000 --> 00:00:11,000

Welcome to this first podcast for 2019. It's real y great that you could be here. Today I'm talking with Jofish Kaye. He's a Principal Research Scientist at Mozil a. And before this, he worked at Yahoo and at Nokia. And what's interesting is that Jofish made a really deliberate decision when he finished his PhD



12

00:00:11,000 --> 00:00:12,000

that he didn't want to pursue an academic career, and instead wanted to work in industry. And it's interesting to hear how he approached this decision making and how his criteria for the different jobs that he's moved into have changed and evolved over time, starting primarily about the people he could work with to becoming much more values driven, and real y being concerned with being able to make an impact in what he does. In fact, there's a strong sense of values and impact in a lot of what he talks about in different ways. So he discusses a lot of his experiences more generally about working in an industry context. And also as he's now moving into more management and leadership roles. So again, as usual, lots of really interesting things here. I hope you enjoy this conversation with Jofish Kaye.



13

00:00:12,000 --> 00:00:13,000

Geri:
01:40


14

00:00:13,000 --> 00:00:14,000

So Jofish, thanks for joining me. [Jofish: My pleasure.] I wanted to talk to you because you made a decision after your PhD to work in industry. I just think it'd be really interesting to hear about your experiences, some part of that decision making process and some of your experiences in playing out.



15

00:00:14,000 --> 00:00:15,000

So where did you do your PhD?



16

00:00:15,000 --> 00:00:16,000

Jofish:
02:00


17

00:00:16,000 --> 00:00:17,000

I did my PhD at Cornel , with Phoebe Sengers. And I was getting to the end of that, and had to make a decision about what to do next. And I think I originally thought that I might go into industry for a few years and then move into academia.



18

00:00:17,000 --> 00:00:18,000

-



19

00:00:18,000 --> 00:00:19,000

1 -



20

00:00:19,000 --> 00:00:20,000

Transcribed by https://otter.ai



21

00:00:20,000 --> 00:00:21,000

Jofish:
02:19


22

00:00:21,000 --> 00:00:22,000

And I think I quite quickly realized that I didn't want to do that anymore, that I was real y happy being in industry. And I was able to do all the things I wanted to do. And I guess it's been whatever, 10 or 12



23

00:00:22,000 --> 00:00:23,000

years now. And I've had a lovely time. [Geri: And you're stil there.] I'm stil there, I'd recommend it.



24

00:00:23,000 --> 00:00:24,000

Geri:
02:36


25

00:00:24,000 --> 00:00:25,000

So did you go and work in industry straight away as your first preference for a job, or you thought you were just trying it out for a bit?



26

00:00:25,000 --> 00:00:26,000

Jofish:
02:45


27

00:00:26,000 --> 00:00:27,000

Absolutely, absolutely. So when I finished up my PhD, I started talking to people and talking to friends, and in particular two interesting opportunities came up at CHI that year. I went and I talked to Elizabeth Churchil and Elizabeth said, "Oh, wel , we're hiring in my group at Yahoo." Great. And I talked to my friend Wendy Ju. And Wendy said, "My friend Tico, over at Nokia, his group is hiring." So I talked to Tico. And they both say "Great, why don't you come out for an interview." And real y the choice there was about - I met some great people in both places. It was a quality of life thing. At Yahoo I was offered Great America, it meant driving everywhere. Whereas Nokia's lab was in Palo Alto, we could walk over to California [?] out for lunch, which we did regularly. Both had great teams and great people. So I went to Nokia. I knew that moving into industry, one of the things I was going to miss was teaching. I've always enjoyed teaching. I've always enjoyed that relationship with students. And I pretty quickly started coaching for Pam Hinds' class.



28

00:00:27,000 --> 00:00:28,000

Geri:
02:51


29

00:00:28,000 --> 00:00:29,000

Did you approach her about that? Or how'd that happen?



30

00:00:29,000 --> 00:00:30,000

Jofish:
03:24


31

00:00:30,000 --> 00:00:31,000

How did that happen? I think it was actual y through her co-teacher, Julie Stanford. [questioningly] Now, how did that happen? Because, no, I met Julie in that class. I don't remember. I think Pam must have reached out to me. But it's possible because Nokia was right next to Stanford, we had a lot of good relationships, right? So I started coaching that, and then Terry Winograd was looking for someone to help coach his class with Josh Cohen on designing liberation technologies. So I coached that for a couple of years. And then I realized that was one of the most fun things I was doing. And so I said, Hey, can I come and teach this with you and Josh? So Terry, and Josh Cohen and myself and a guy called Zia Yusuf, who had just left SAP at the time, we taught that class for a couple of years. And then Stanford's HCI Department said, Hey, can you teach the intermediate HCI Class? And so I taught that, me and John Tang, because they said, Would you teach this? And I said, Only if I get to pick my co-teacher, and they were like, Okay. And I'd wanted to know John better, I knew him a bit, but not that wel . So he and I taught it. And then they were like, Okay, we need a few more people to teach this, having two people from outside the university. And so Michael took it back over and changed it more into a model with five or six external industry people. And maybe I should back up a little bit. I always feel universities have things that they do particularly wel . And I always feel like Stanford's secret sauce is being able to bring in people from industry. Because it's in the middle of Silicon Valley. And it makes such a difference. And that's what they do really, really well. The great thing you get from a Stanford



32

00:00:31,000 --> 00:00:32,000

-



33

00:00:32,000 --> 00:00:33,000

2 -



34

00:00:33,000 --> 00:00:34,000

Transcribed by https://otter.ai



35

00:00:34,000 --> 00:00:35,000

education, is that exposure to all these people who are really doing stuff in the real world, and other schools have other wonderful capabilities. But that's Stanford's sort of special sauce.



36

00:00:35,000 --> 00:00:36,000

Geri:
06:10


37

00:00:36,000 --> 00:00:37,000

Absolutely. Did you have any trouble negotiating that with Nokia?



38

00:00:37,000 --> 00:00:38,000

Jofish:
06:14


39

00:00:38,000 --> 00:00:39,000

I had a fabulous manager, Mirjana Spasojevic. And she was total y happy with that. I later switched jobs, and again, I had managers who were enthusiastic about this, and saw the value that it brought to the corporation.



40

00:00:39,000 --> 00:00:40,000

Geri:
06:31


41

00:00:40,000 --> 00:00:41,000

In what way? How did you argue that?



42

00:00:41,000 --> 00:00:42,000

Jofish:
06:33


43

00:00:42,000 --> 00:00:43,000

It brings visibility. In al of these cases -- I was at Nokia for a few years, I was at Yahoo, I'd been at Mozil a -- in none of these cases were we the biggest dog in the room, right. And yet, we could show up and be present in the minds of these really top undergraduates, these top graduate students. It was a way to have that visibility to have that influence. And to tie in these researchers into the stuff that we were doing on a day to day basis. And we can point to a couple of papers that came out that we were able to do because we had a really healthy relationship with Stanford and with Stanford students, that I think if we hadn't had that larger context of a university relationship would be much harder to do.



44

00:00:43,000 --> 00:00:44,000

Geri:
07:17


45

00:00:44,000 --> 00:00:45,000

Yeah. Was that also enabling you to identify good students to possibly recruit or was that less?



46

00:00:45,000 --> 00:00:46,000

Jofish:
07:24


47

00:00:46,000 --> 00:00:47,000

A hundred percent. Recruiting, internships, right, those two tie in closely together.



48

00:00:47,000 --> 00:00:48,000

Geri:
07:30


49

00:00:48,000 --> 00:00:49,000

I real y liked the fact that you've identified, whichever way the first relationship with Pam Hinds came about, the fact that you said that was something you really loved. And then making a way for that to happen, and then to be a regular part of your ongoing professional life.



50

00:00:49,000 --> 00:00:50,000

Jofish:
07:48


51

00:00:50,000 --> 00:00:51,000

And I kept it up. I try to do one quarter a year, that was sort of how it ended up working. And I do little things here and there. But everybody does, you know, "Can you guest teach this class?" of course, right. But those one quarter a year of actual teaching. That's something I was able to fit into my professional career, up until the time I was chairing CHI. And that's when I was like, Alright, I can't do both those two things. So I stopped it then. And now, since I moved to Mozil a, I have more responsibilities. There's a lot of extra work, and I just don't have the time, right?



52

00:00:51,000 --> 00:00:52,000

Geri:
08:28


53

00:00:52,000 --> 00:00:53,000

Do you miss it? Or does it feel like you've done it for a while and it's fine?



54

00:00:53,000 --> 00:00:54,000

-



55

00:00:54,000 --> 00:00:55,000

3 -



56

00:00:55,000 --> 00:00:56,000

Transcribed by https://otter.ai



57

00:00:56,000 --> 00:00:57,000

Jofish:
08:35


58

00:00:57,000 --> 00:00:58,000

Maybe both of those at the same time? I do miss the excitement. I don't miss the logistics.



59

00:00:58,000 --> 00:00:59,000

Geri:
08:47


60

00:00:59,000 --> 00:01:00,000

Does that mean the administration around the course?



61

00:01:00,000 --> 00:01:01,000

Jofish:
08:49


62

00:01:01,000 --> 00:01:02,000

It was the administration, but more those things like parking on campus and having to think about parking on campus. Those sort of niggly things end up making such a difference. I love the fact that right now, for example, I have minimal commute, it takes me about eight minutes to bike from my home to my office.



63

00:01:02,000 --> 00:01:03,000

Geri:
09:09


64

00:01:03,000 --> 00:01:04,000

Bike, not even not even drive. Yeah. That's hardly worth getting on the bike for.



65

00:01:04,000 --> 00:01:05,000

Jofish:
09:15


66

00:01:05,000 --> 00:01:06,000

Absolutely, right? But then to say, Wel , okay, now I'm going to take the half hour to go over to Stanford, and then half hour to come back. That's a lot of extra time. Right?



67

00:01:06,000 --> 00:01:07,000

Geri:
09:24


68

00:01:07,000 --> 00:01:08,000

Yeah. Sounds like it was good for a while while it happened. And time has moved on. [Jofish: Absolutely. Yes.] So you said, did you apply? I'm just interested about did you apply for faculty positions at all?



69

00:01:08,000 --> 00:01:09,000

Jofish:
09:37


70

00:01:09,000 --> 00:01:10,000

No, none.



71

00:01:10,000 --> 00:01:11,000

Geri:
09:39


72

00:01:11,000 --> 00:01:12,000

What was it about an academic role that you didn't connect with or want to do?



73

00:01:12,000 --> 00:01:13,000

Jofish:
09:46


74

00:01:13,000 --> 00:01:14,000

Applying for grants. Particularly you look at the sort of the global funding situation for tertiary education, and I'm looking at these funding situations where people are getting, you know, 7% of grants are getting funded, 5%.



75

00:01:14,000 --> 00:01:15,000

Geri:
10:02


76

00:01:15,000 --> 00:01:16,000

Even 1%, 2% in some European cal s.



77

00:01:16,000 --> 00:01:17,000

Jofish:
10:06


78

00:01:17,000 --> 00:01:18,000

Absolutely. And I've done my share of being on NSF panels, whatever. And I've looked at these proposals coming in and easily 80% of these are fundable. And yet only 5 or 10, or 20. You end up being so limited in what you can do. And let me present a possibly controversial opinion, I've been able to have a lot of freedom to do what I want, because I am in industry, and because I am not, therefore



79

00:01:18,000 --> 00:01:19,000

-



80

00:01:19,000 --> 00:01:20,000

4 -



81

00:01:20,000 --> 00:01:21,000

Transcribed by https://otter.ai



82

00:01:21,000 --> 00:01:22,000

limited by the need to get funding to be able to support the work that I do. And I think there is a story which is aspirational and valid and a real y good narrative, which is the freedom of academia to do whatever you want. But in a world in which that academia is not suitably publicly funded, and getting external funding is such a key part. Al of a sudden, that doesn't happen anymore. Okay, now, what's going on?



83

00:01:22,000 --> 00:01:23,000

Geri:
11:00


84

00:01:23,000 --> 00:01:24,000

I total y agree. I total y agree, because it's the freedom in quotes, but often it's according to what the funders' agendas are, because increasingly funding agencies are setting the priorities of what they want to fund. So you have to reshape, reinvent, to fit a call. And I know for myself, when I've looked back at my recent publications, a good number of them, or most of them have not come from my funded projects. They've come from more skunkworks projects or projects with students, and other avenues. And even though the funded projects have taken an enormous amount of work -- often that work's been more about reporting and administration and going to meetings and crazy stuff.



85

00:01:24,000 --> 00:01:25,000

Jofish:
11:49


86

00:01:25,000 --> 00:01:26,000

Yeah, there's a lot of overhead there. And it's interesting that now that I'm moving into this position, where at Mozil a, I run our Mozil a research grant process, and we've given out something like probably 30-odd grants over the last two years that I've been running it, we've got a deadline in a few days' time for the fourth round. I took it over and I wanted to move it from being kind of a smal - it was a very focused grant program, right? Like, if you were doing something that I knew that I needed, I'd be like, Oh, here you go. So it wasn't real y a public cal . And we changed it to say, let's have a public cal , let's be much more open about the diverse kinds of research that we're funding. And I've been really delighted by that process.



87

00:01:26,000 --> 00:01:27,000

Geri:
12:37


88

00:01:27,000 --> 00:01:28,000

So what would be your success rate, then if you say, you awarded 30 so far?



89

00:01:28,000 --> 00:01:29,000

Geri:
12:47


90

00:01:29,000 --> 00:01:30,000

Is that more limited by budget?



91

00:01:30,000 --> 00:01:31,000

Jofish:
12:47


92

00:01:31,000 --> 00:01:32,000

Wel , let's do some math. So the first time around, we had, I think 36 applications, and we accepted about 10 or 12 of them, it was 12 of them. [Geri: That's really good.] Second time around, we had 78



93

00:01:32,000 --> 00:01:33,000

people apply, right. Word got out. And again, it was roughly about 10 or a dozen, right.



94

00:01:33,000 --> 00:01:34,000

Jofish:
12:53


95

00:01:34,000 --> 00:01:35,000

It's 100% limited by budget, if I had twice the budget, I would put out twice the things. In the spring, we had 115 people apply. And again, sort of about 10. This fall, we had some budget limitations, where we knew we would only be able to give out a small number of grants. And so instead of doing these big open calls, which are great and really inspirational and exciting, we did a really focused one in which we picked three particular areas and said these are the exact questions we want an answer to.



96

00:01:35,000 --> 00:01:36,000

Geri:
13:51


97

00:01:36,000 --> 00:01:37,000

-



98

00:01:37,000 --> 00:01:38,000

5 -



99

00:01:38,000 --> 00:01:39,000

Transcribed by https://otter.ai



100

00:01:39,000 --> 00:01:40,000

That fitted more business priorities.



101

00:01:40,000 --> 00:01:41,000

Jofish:
13:53


102

00:01:41,000 --> 00:01:42,000

100% aligned with business priorities, where we could say things like -- the one that is closest to my team is about listen-ability. We're doing a lot of work around reading text aloud. Can you listen to the web?, is sort of the story that we're tel ing. So what does a listen-ability score look like? How do we know if text is listen-able or not? And how can we make hard-to-listen-to text easier to listen to? So we've got a specific cal around that. But that's a much more focused process. Partly, my hope is to actually reduce the number of applications by having it more focused. The other thing that happened was a change in how we started to think about these grants. Because when you have seven people apply and you give six people funding or something like that, then the way that you have impact through your cal is by making sure that those seven or eight people you give the funding to are able to do what they need to do. If you're in this opposite situation in which you're only able to fund a small percentage, then the ways you can generate value for the corporation and for the people submitting grants, you've got to sort of rethink that. And so I tried to make it so that we could find ways to collaborate with these people. Maybe we don't give them money, but perhaps we can give them some data that's really exciting, right? And more and more thinking about Mozil a's role, we have a strong corporate mission, right? We really believe in the web and keeping it open and accessible to all. If we can work with someone and say, Look, we weren't able to fund your work. But it sounds like you might find this dataset that we've got real y interesting. So we have a data set of people who have explicitly opted in to have all of their browsers' stuff tracked. I want to be real y clear about how much they have deliberately opted in that they really wanted to do this, right. And so we have that dataset, the pioneer data set that we can share with people.



103

00:01:42,000 --> 00:01:43,000

Geri:
13:55


104

00:01:43,000 --> 00:01:44,000

Which is invaluable for researchers, because getting access to that sort of data on your own can be real y difficult, especial y after recent events.



105

00:01:44,000 --> 00:01:45,000

Jofish:
16:04


106

00:01:45,000 --> 00:01:46,000

Absolutely. Another lovely example of this: We did a survey through the Mozil a Foundation about what IoT devices people had, but we put it out through Firefox. And we kept it out there for quite some time.



107

00:01:46,000 --> 00:01:47,000

And we ended up gathering, over the course of a month, about 184,000 survey respondents. And that's a fascinating data set. So I then worked with our trust and security people and they wouldn't let me put out the qualitative work, but they let me put out the quantitative numbers. And the metadata, right? Like, what country was it coming in? Right, what time of day, things like this, right? No IP addresses, nothing personally identifiable. So we put out that dataset, Creative Commons licensed it. And then Jed Brubaker at CU Boulder, took that dataset and used it as part of his intro class. And so they had whatever it was, 80 freshman, who had a unit, which was take this very large dataset and start to work with it, right? It changes the ways that you do work if you've got 184,000 replies instead of 10 or 100.



108

00:01:47,000 --> 00:01:48,000

How can you form meaningful questions? What are the tools you need to do that? And I'm real y delighted to see that happen. I'd love to see that happen more.



109

00:01:48,000 --> 00:01:49,000

Geri:
17:22


110

00:01:49,000 --> 00:01:50,000

-



111

00:01:50,000 --> 00:01:51,000

6 -



112

00:01:51,000 --> 00:01:52,000

Transcribed by https://otter.ai



113

00:01:52,000 --> 00:01:53,000

So you're facilitating al sorts of research, even apart from directly funding. That, as you said, is in line with the mission of Mozil a around the open web. And that also seems like good strategies for keeping Mozil a front of mind for people, or putting it in front of people more often.



114

00:01:53,000 --> 00:01:54,000

Jofish:
17:44


115

00:01:54,000 --> 00:01:55,000

I hope so. I mean, I do believe in the mission, right? And I believe the value of an open ecosystem, the great thing about the web is twofold. Anyone can publish anything on it, you don't have to ask a large company, if it's okay, "Can I publish something on your platform", anyone can publish, and anyone can read it from any device. It's not like you have to have special devices that are blessed by one company to run the web. Anyone can do it. And that's the thing that runs through all the work we do. So we're doing a bunch of IoT stuff now.



116

00:01:55,000 --> 00:01:56,000

Geri:
18:18


117

00:01:56,000 --> 00:01:57,000

IoT for people who don't know is Internet of Things, is about al sorts of devices that are enabled to connect to the internet and share data and talk to other devices.



118

00:01:57,000 --> 00:01:58,000

Jofish:
18:31


119

00:01:58,000 --> 00:01:59,000

Precisely. And our take on it, we're cal ing web of things, which is about how can we take this device that might be a smart light bulb, but it's a smart light bulb that's run by, owned by, made by Philips or Samsung? And we're like, wel , let's give that smart light bulb a URL, with security, with ways to access it, not just an open URL on the internet. But the idea that then any device can interact with that light bulb, right? And that's sort of the vision behind the web of things stuff. And it's a very sort of Mozil a story. How can we take this global public resource that's open to all and facilitate?



120

00:01:59,000 --> 00:02:00,000

Geri:
19:08


121

00:02:00,000 --> 00:02:01,000

Yeah. That sounds real y interesting.



122

00:02:01,000 --> 00:02:02,000

Jofish:
19:11


123

00:02:02,000 --> 00:02:03,000

I think so. Not to report the company line, right. Yeah.



124

00:02:03,000 --> 00:02:04,000

Geri:
19:16


125

00:02:04,000 --> 00:02:05,000

I can see that you're real y enthusiastic, you know, and then your body language and the just the sort of, you know, the eyes lit up, and that that's something that you really care about. So you've worked at Nokia, then Yahoo. You eventual y got to Yahoo. And then Mozil a? What were the triggers for making those shifts?



126

00:02:05,000 --> 00:02:06,000

Jofish:
19:38


127

00:02:06,000 --> 00:02:07,000

So Nokia closed that lab, or closed most of that lab, so we got laid off. That was the first one, was a necessity. The second one, Yahoo closed down the vast majority of Yahoo labs. And I was not laid off but a lot of the people I was working with were. I stayed for about another six months after a lot of my colleagues were laid off. And I did some super work during that time, I really enjoyed what I did. I did some work with some people in the accessibility group about doing better keyboards for blind people. I did some interesting work around diversity and inclusion across the workforce and things like that.



128

00:02:07,000 --> 00:02:08,000

-



129

00:02:08,000 --> 00:02:09,000

7 -



130

00:02:09,000 --> 00:02:10,000

Transcribed by https://otter.ai



131

00:02:10,000 --> 00:02:11,000

Geri:
20:29


132

00:02:11,000 --> 00:02:12,000

And were they driven by your interests? Or how did they come about?



133

00:02:12,000 --> 00:02:13,000

Jofish:
20:33


134

00:02:13,000 --> 00:02:14,000

Those were I being made principal scientist, and I thought to myself, what are the things that I should do as principal scientist. Particularly when you're in that mode, where you are less concerned about what the company thinks of you. And so I had this sort of six months of doing great work and having great fun. I was on the membership team. And I think I made some real contributions there, as well, that were much more focused on what the immediate problems at the company were. But it got to the end of that and I was like, Alright, it's time to be done. But in fact, the thing that I started to realize, much in the same way, I had gone back and thought about the teaching and gone to Terry and said, This is the thing I've enjoyed the most. I thought about what I had enjoyed the most in the last couple of years of my career. I had a very supportive manager, Ayman Shamma. When I'd gone in to chairing CHI, I had so much fun -



135

00:02:14,000 --> 00:02:15,000

Geri:
21:25


136

00:02:15,000 --> 00:02:16,000

and CHI is a major conference in our research area -



137

00:02:16,000 --> 00:02:17,000

Jofish:
21:29


138

00:02:17,000 --> 00:02:18,000

It was about 4000 people. I know that you're foolish enough to take it on, doing it next time. But I had been working on it with Al ison Druin, who was at the University of Maryland at the time, and is now at the Pratt Institute. We'd come up with a theme for the conference, which was "CHI for good". And CHI for good was fine. We put it on the logos, whatever. But it led us to think about what we were trying to do. And it was great, because we were trying to answer questions like, well, what's the CHI for good answer? And I love this idea of CHI for good that you can do human computer interaction, for the sake of the public good, not just for the sake of the bottom line, or to show value in some sort of economic way. But real y, this is great, right? So when I thought about things I had done, CHI for good was on the top. And so I made a list of the things I wanted to be doing in my next position. And the first one was CHI for good. And then I went around to talk to my friends, and was chatting to people, "I'm real y looking for a place where I can do CHI for good". And I was actually playing guitar over at a friend's place with this guy, Sean. And we're playing guitar and various other instruments that none of us play particularly well. And I say I'm looking for this CHI for good. And he's like, "Well, do you want to come and talk to me about Mozil a?" And so I came to talk to him. And I hadn't really thought about Mozil a as an option. But I was really impressed by what I heard. I called up another friend who was working there.



139

00:02:18,000 --> 00:02:19,000

And it's a fascinating organization. I was like, yeah, this is a great place to be, I'm real y excited by this.



140

00:02:19,000 --> 00:02:20,000

Jofish:
21:38


141

00:02:20,000 --> 00:02:21,000

I think I'm the only person on the planet who likes job searches. Because you get to reinvent yourself.



142

00:02:21,000 --> 00:02:22,000

[Geri: That's an interesting way of putting it.] I don't know how true this is in academia. But I feel like in industry, let's say you do A, B, C, D and E, right? And in fact, you're only evaluated on A and B, right?



143

00:02:22,000 --> 00:02:23,000

You do C and D because they've somehow become your responsibility. But the only things you're evaluated on are A and B. Well, when you switch jobs, you can say, Well, I've been doing A, B, C, D



144

00:02:23,000 --> 00:02:24,000

and E, I want to be a person who does, say, B and E. Then you sort of rewrite your resume, like "I, a person who does B and E, should do this." And then you go to people and they're like, "We've been



145

00:02:24,000 --> 00:02:25,000

-



146

00:02:25,000 --> 00:02:26,000

8 -



147

00:02:26,000 --> 00:02:27,000

Transcribed by https://otter.ai



148

00:02:27,000 --> 00:02:28,000

looking for someone to do B and E, we're looking for an E person who really understands B!". And I love that you get to reinvent who you are, and reinvent what your priorities are. And so it's like, well, my priorities of CHI for good has gone from being, you know, E, to something that I want to be A or B, right.



149

00:02:28,000 --> 00:02:29,000

Geri:
23:25


150

00:02:29,000 --> 00:02:30,000

And you can say, as an added bonus, you get the steak knives of A, C and D!



151

00:02:30,000 --> 00:02:31,000

Jofish:
24:07


152

00:02:31,000 --> 00:02:32,000

Absolutely. Wel , it's so useful that you know something about A, C and D right, you know, Geri 24:30



153

00:02:32,000 --> 00:02:33,000

Do we do that in academia? I guess we do, actual y, to some extent, but it's maybe harder to pul off.



154

00:02:33,000 --> 00:02:34,000

Yes, because our job ads for academic positions wil say they want someone with particular expertise that you might have done in a smal way. So you try to sort of play that up and foreground that more, but I guess if you don't have the publications to back you up, it is a harder case to make. Al of these positions in industry have been research positions rather than in the product delivery sort of side.



155

00:02:34,000 --> 00:02:35,000

Jofish:
25:11


156

00:02:35,000 --> 00:02:36,000

That's true. The Mozil a organization that I'm in now is real y transitioning. It is now the R&D org, but it is not really a research org in the traditional sense, right? In particular, we have the onus opportunity to ship product directly to customers, it includes some very researchy things. So I have colleagues who work on Rust, which is a programming language that's a replacement for C, that's much safer than C, and people who use it think it's much better. We're never going to make money off Rust, the intent is not to make money off Rust. And in fact, explicitly the intent is to make it so that Rust is its own. We don't call it Mozil a Rust, it is Rust, and it is run out of its own sort of community and thing. And that's great. But that's part of the organization that I mean. But really, the focus over the last year or two has switched. And we've really been pushing to how can we make products, right, and how can we make products as a way to have impact in the world in a sustainable way?



157

00:02:36,000 --> 00:02:37,000

Geri:
26:16


158

00:02:37,000 --> 00:02:38,000

Are you okay with that shift? Because having done more research, you have an extensive publication profile through all those industries. Does that mean you're publishing less now? Or do you see in the future publishing less?



159

00:02:38,000 --> 00:02:39,000

Jofish:
26:32


160

00:02:39,000 --> 00:02:40,000

That's an interesting question. So I'm entirely okay with it. I think it's a real y exciting way to have more impact in the world. I am really concerned about the way that we treat publications as THE way to make success in the world. Right. And I love what Casey Fiesler has done with this year's CSCW conference, where she's encouraged everyone to do a Medium post, to present their work in an open way to the public, to people who aren't going to download a paper from an academic institution. I think that's so important. And in our Mozilla research grant funding, we have a line in there that says, basically send us any papers that you write, and we encourage you to do some publicly accessible way to do this.



161

00:02:40,000 --> 00:02:41,000

Because this ties back to this previous question about university funding. I think it is so important and so incumbent upon research as a field to make clear and visible and important, how valuable what is



162

00:02:41,000 --> 00:02:42,000

-



163

00:02:42,000 --> 00:02:43,000

9 -



164

00:02:43,000 --> 00:02:44,000

Transcribed by https://otter.ai



165

00:02:44,000 --> 00:02:45,000

we do? And the more we have this impression as being a bunch of academics stuck in an ivory tower doing stuff, you know, we're just talking to each other. I see no value in that. Right. We need to be taking seriously this cal for public outreach.



166

00:02:45,000 --> 00:02:46,000

Geri:
27:51


167

00:02:46,000 --> 00:02:47,000

So you're very clear, then that while you do have publications previously, what you're real y driven by, what real y underpins it al is that desire to have an impact. That's also reflected in the CHI for good. It seems like the CHI for Good is a refinement of that wanting to make a difference, have an impact, agenda. And the implications of it for where you are now might be, there is more product delivery, but it doesn't sound like it's constraining in the way that sometimes you hear people working in industry contexts, feeling constrained by having to deliver to the company's strategic direction.



168

00:02:47,000 --> 00:02:48,000

Jofish:
28:27


169

00:02:48,000 --> 00:02:49,000

I think one of the things that real y helps, is -- obviously there's two things. One, Sean, my boss is very, it's up to us to figure out what we're doing and then do it. There's not sort of this being handed down from up above, "You must deliver this thing."



170

00:02:49,000 --> 00:02:50,000

Geri:
28:44


171

00:02:50,000 --> 00:02:51,000

To this product, or this particular product.



172

00:02:51,000 --> 00:02:52,000

Jofish:
28:47


173

00:02:52,000 --> 00:02:53,000

And the flip side is also that you get to tel the stories that you need to tel in the ways you tel them.



174

00:02:53,000 --> 00:02:54,000

And sometimes that's shipping some piece of product, sometimes it's doing a blog post, sometimes it's sharing a dataset. And those are all ways to have impact in the world. And it's tricky to measure. One of the advantages of Mozil a, and I'm sorry to keep coming back to the company line. We have one shareholder, right? I work for the Mozil a Corporation. It has one shareholder, which is the Mozil a Foundation. And the Mozil a Foundation is a Not For Profit Organization, right? It's a charity. And that means that we don't have to return shareholder value in the short term. And that frees you to do these things that are a little bit longer, that have a little bit more space. And there's a double bottom line, right? You've got to make profit at the end of the day because Geri 29:50



175

00:02:54,000 --> 00:02:55,000

The company's got to keep surviving.



176

00:02:55,000 --> 00:02:56,000

Jofish:
29:51


177

00:02:56,000 --> 00:02:57,000

Absolutely. And that's total y crucial. At the same time, it means there are things that we wouldn't do.



178

00:02:57,000 --> 00:02:58,000

Yes, we could take al the data and then sel it to people. That's so far away from what we would possibly do. And for other companies that aren't so mission driven, and are shareholder driven in particular, if that produces short term value, then your fiduciary duties can mean that that is what you must do, if that's the way to do it. And to try and move away from this notion of having to just produce short term profit. It's remarkably empowering. I feel very lucky. I want to be really clear about that.



179

00:02:58,000 --> 00:02:59,000

Geri:
30:33


180

00:02:59,000 --> 00:03:00,000

-



181

00:03:00,000 --> 00:03:01,000

10 -



182

00:03:01,000 --> 00:03:02,000

Transcribed by https://otter.ai



183

00:03:02,000 --> 00:03:03,000

But it also says that it's worth putting in some time to think as you talked about before, what's important to you? What are your values? And where is there an alignment and a fit, where you can real y feel like you can bring together the power of the company and your passions and interests to make a bigger impact and a difference?



184

00:03:03,000 --> 00:03:04,000

Jofish:
30:55


185

00:03:04,000 --> 00:03:05,000

Absolutely. And for this, I think the ways that I've thought about my job search have changed over time.



186

00:03:05,000 --> 00:03:06,000

I went through stages where the group of people I was working with was absolutely key, right? Did I know those people, had I met them, did I feel confident about working with those people. With Mozil a, I didn't know the people I was coming to work with, with the exception of Sean, who I'd worked with at Nokia, and someone I knew from undergrad who was far off in Boston somewhere. But I had sort of ended up with this values driven decision making. And I had other stuff as well, I had a list of I think it was six or seven things, which the first one was CHI for good. I wanted to be in a management role. I wanted to be doing something more strategic and not UX, right? I wanted to do HCI not UX. For me, that was the right decision. It wouldn't be for other people. But for me, that was the right decision.



187

00:03:06,000 --> 00:03:07,000

Things like I love being able to talk about my work. I have some very good friends at Apple, for example, who can't talk about their work at al . And I love the transparency of an organization like Mozil a where I can go and talk about my stuff. I don't have to make you sign an NDA before I talk about what I'm doing. And it's very empowering and it reduces the amount of friction in the system. And every so often I'l get asked to sign an NDA. And it's a big important thing. Of course we do. But the default is that we don't. And when someone says, "Can you sign this NDA?", we point them at our lawyers whose job it is to say, "We don't really sign NDAs." And generally, you don't have to, it turns out.



188

00:03:07,000 --> 00:03:08,000

Geri:
32:31


189

00:03:08,000 --> 00:03:09,000

What was your process for coming up with those points? Did you explicitly sit down give yourself a bubble of time, piece of paper, write them down? Or was it just something that mulled around in your head?



190

00:03:09,000 --> 00:03:10,000

Jofish:
32:42


191

00:03:10,000 --> 00:03:11,000

Somewhere between the two. I ended up with six or seven of them. I wrote down some of them, scribbled them down on a bit of paper. Those three that I just mentioned. I talked them over with my wife, which was a hugely important part of this process. Location was very important. I told you about the eight minutes. I didn't have "must be able to bike there in eight minutes", but I knew I didn't want -- I live down in Mountain View, which is due south of San Francisco by about an hour. I knew I didn't want to have an hour commute each way. If I had half an hour or 45 minute bike ride that's different. That's okay. So location was important. I put salary on there. Because I didn't want to give this list to a recruiter and be "Oh, and I don't care about how much I get paid." That seemed like a poor strategic decision. So I put that on the list. The honest answer is it wasn't a huge motivator. But it's definitely an issue. Right. You know, there's a family?



192

00:03:11,000 --> 00:03:12,000

Geri:
33:43


193

00:03:12,000 --> 00:03:13,000

Yeah, you've got a family to support. So there's a responsibility and just valuing your own worth and experience that you bring. It sounds like you can tick all your boxes with this.



194

00:03:13,000 --> 00:03:14,000

-



195

00:03:14,000 --> 00:03:15,000

11 -



196

00:03:15,000 --> 00:03:16,000

Transcribed by https://otter.ai



197

00:03:16,000 --> 00:03:17,000

Jofish:
33:56


198

00:03:17,000 --> 00:03:18,000

I think so. I mean, I'm remarkably lucky. And I definitely acknowledged the luck, right? I mean, I'm presenting it here as "I made a great plan, and I executed it." And there's some of that, but …



199

00:03:18,000 --> 00:03:19,000

Geri:
34:10


200

00:03:19,000 --> 00:03:20,000

I think that's a common theme with so many people I've talked to. Very few people have sat down and had this strategic plan that they've made happen. It's been happenstance and coincidental meetings, playing the guitar with someone who happens to say, come and chat and, you know, it's just being in situations and being open to, you know, I guess, recognize those opportunities and then follow them up.



201

00:03:20,000 --> 00:03:21,000

Jofish:
34:34


202

00:03:21,000 --> 00:03:22,000

Yeah, it's tricky within the academic world. Partly it's a supply and demand question, right. I look at what a good student wil be doing. I'm fil ing out recommendation letters for one of my older interns who's bril iant, and I look at that list of recommendation letters and it's daunting. And you're like, Well, this is this is such a long list because it's so hard to figure out where you are. I don't see a better solution to this system.



203

00:03:22,000 --> 00:03:23,000

Geri:
35:10


204

00:03:23,000 --> 00:03:24,000

I think the US system seems particularly competitive and difficult. And I know that it's getting increasingly that case everywhere. And then other sectors, there's more of the casualization of the postdoctoral workforce where it's harder to get into more tenure track positions. So there are different challenges, but challenges in all countries, I think.



205

00:03:24,000 --> 00:03:25,000

Jofish:
35:31


206

00:03:25,000 --> 00:03:26,000

There's value, I think in recognizing it as an issue rather than, "Wel , this is the way it is, and it has always been, and should always be." I'm going to throw something out here I haven't thought about at all. It's always dangerous. In the UK, when you apply to universities, you fil out one form, it's called



207

00:03:26,000 --> 00:03:27,000

[Bacus?]. And you fil out the UCAS form. And I think you can fit 12 --



208

00:03:27,000 --> 00:03:28,000

Geri:
35:55


209

00:03:28,000 --> 00:03:29,000

You mean as a student?



210

00:03:29,000 --> 00:03:30,000

Jofish:
35:57


211

00:03:30,000 --> 00:03:31,000

As a student, sorry, yes. To be explicit: this is applying to be an undergraduate. You fil out one form, and you write one essay, and you apply to 12 universities, along with your transcripts, whatever. In the US, you apply to each one university separately, there is a thing called the Common Application form, but most universities don't use it. It's incredibly complicated and very expensive. If you apply to you know, a dozen universities, that might be $100 per university or more.



212

00:03:31,000 --> 00:03:32,000

Geri:
36:25


213

00:03:32,000 --> 00:03:33,000

You pay to apply? [Jofish: Yes. Yeah!] That real y limits access. If we think about equity…



214

00:03:33,000 --> 00:03:34,000

-



215

00:03:34,000 --> 00:03:35,000

12 -



216

00:03:35,000 --> 00:03:36,000

Transcribed by https://otter.ai



217

00:03:36,000 --> 00:03:37,000

Jofish:
36:35


218

00:03:37,000 --> 00:03:38,000

Absolutely, and I think some universities are doing better about, you know, waiving application fees, there is this common application form for some cases. But it doesn't delight me as a process that you look at and…



219

00:03:38,000 --> 00:03:39,000

Geri:
36:54


220

00:03:39,000 --> 00:03:40,000

Yeah, and I think there are lots of processes in academia and universities that we need almost sort of go back to the drawing board and start again on and say, Do we really need this? And who is it serving? Whose needs is it serving?



221

00:03:40,000 --> 00:03:41,000

Jofish:
37:08


222

00:03:41,000 --> 00:03:42,000

And whose needs are serving? And who ends up paying for that? And I mean, I think there's a bigger question. I was just talking with Leysia Palen, who's chair of the informatics department at CU Boulder in Colorado. What does the future of the public university look like? I happen to like universities, I think they're real y real y important. And they're very good for the long term survival of al that is good in the world, kind of thing. What does it mean to lean into that in a useful way? Right. This is something that she's actually directly concerned with as someone who runs a whole department. What does it mean to keep a public university going, right? To loop back to our question about funding changes, right. And if this is no longer seen as a priority by governments, you look at sort of successive years of cutting funding, I believe just about everywhere. [Geri: Everywhere, yeah.] I see the potential for universities to reduce inequality at a societal level in a fundamental way. To me, that seems like a pretty good thing. I don't like Gini coefficients being big right? I want to see them a little bit smaller. How we engage with that as people within a public sphere, I don't know what the right answer is.



223

00:03:42,000 --> 00:03:43,000

Geri:
38:32


224

00:03:43,000 --> 00:03:44,000

Just a couple of things about your role. So how many hours a week do you work? What's the sort of working pattern?



225

00:03:44,000 --> 00:03:45,000

Jofish:
38:43


226

00:03:45,000 --> 00:03:46,000

It's a good question. I drop the kids off at school. I have three children. So I look like this. Two of them are seven years old. They go to the local public project based learning school. So I drop them off by 8.30. And then my little one, Wolfie, he's four years old, so I drop him off at preschool. Then I come home and drop the car and bike to work. And I'm usually there by nine-ish. I try not to have 9am meetings because it's just so tight, before drop-off I have to read two books to Wolfie. So I don't try and have meetings before 930 or 10. And then I try and be home by six because I want to see the kids. And I don't real y work weekends, he says on a -- is it a Sunday? It's a Tuesday now. Conferences are sort of weird, right. But it's not a routine thing. Routinely, I do not work weekends [Geri: Or evenings?] Very little. Before the kids were born, I used to. I had entire research projects that I did after dinner, right, sitting on the couch. I'm so tired. Like, I'm so tired by the time that we've had dinner, I've put the kids to bed, I've read them a chapter of whatever we're reading right now. It gets to nine o'clock, all I want to do is like, let's watch something short and funny that you know, is amusing. And then go to sleep. So yeah, it's something like eight or nine hours a day, five days a week.



227

00:03:46,000 --> 00:03:47,000

Geri:
40:39


228

00:03:47,000 --> 00:03:48,000

-



229

00:03:48,000 --> 00:03:49,000

13 -



230

00:03:49,000 --> 00:03:50,000

Transcribed by https://otter.ai



231

00:03:50,000 --> 00:03:51,000

That's fine. And does that feel like it's, are you pressured to try to keep doing that, like from other work?



232

00:03:51,000 --> 00:03:52,000

Or is it something that you're able to manage? And make work?



233

00:03:52,000 --> 00:03:53,000

Jofish:
40:50


234

00:03:53,000 --> 00:03:54,000

It works pretty wel . I would say, I work from home, maybe probably not quite a day a week anymore, I end up with more meetings than I want. But there's no particular need to go into work if I don't need to.



235

00:03:54,000 --> 00:03:55,000

But they do feed me there, which is nice. There's definitely been times when I'm like, Well, I've worked from home al morning, I have to go for lunch anyway. I might as wel go to work [Geri: Eight minutes cycle into it.] I did have one moment I remember I needed to fax something. And I started to look at installing some fax driver, and I decided it'd be less time to go to work, put it into a fax machine and then come home than actual y get some horrible driver to work on my machine.



236

00:03:55,000 --> 00:03:56,000

Geri:
41:36


237

00:03:56,000 --> 00:03:57,000

So you've moved into a management role as wel ? And what's that experience been like? What are your lessons learned and strategies for doing that.



238

00:03:57,000 --> 00:03:58,000

Jofish:
41:46


239

00:03:58,000 --> 00:03:59,000

So insert some disclaimer about who knows how good I am at this. So I have a wonderful team. So hire total y awesome people. That's what I did.



240

00:03:59,000 --> 00:04:00,000

Geri:
41:58


241

00:04:00,000 --> 00:04:01,000

So they were al new, al people that you hired?



242

00:04:01,000 --> 00:04:02,000

Jofish:
42:02


243

00:04:02,000 --> 00:04:03,000

I brought them in. One of them was already with the company, just sort of in a different role. She's my engineer, Tamra. She's based in Seattle now, she has been with Mozil a a bit longer than I have.



244

00:04:03,000 --> 00:04:04,000

Highly, highly competent. And a lot of what I'm able to do is because she's able to do things like run contractors, and I never see that actual part of that management process, right. So I have an engineer, I have a designer cal ed Abe. It took forever to hire someone because I was looking for this sort of unicorn that would be able to do a whole bunch of different design stuff. And I had the HR recruiting people sitting down with me and saying, "Look, you're not going to get your unicorn, you need to cut down." And then Abe showed up. So we've got Abe and then we've got Janice Tsai. She was at Microsoft before this. She's got a PhD from CMU with Lorrie Cranor, real y top tier researcher, real y able to turn around stuff very, very quickly. She's got a deep understanding of the field. She's at the conference. There's something nice, Sean, my boss has a sort of Noah's Ark theory, which is that you want people to have someone to talk to who shares and, he'd use this wording, who shares the same epistemological assumptions about knowledge in the world. And Janice, it's great to have another HCI person to work with, because I can say, "Why don't we do a Wizard of Oz study on this? See how it goes? And if that doesn't work, we'l throw it on Mechanical Turk." And I don't have to explain words there. It just works, right?



245

00:04:04,000 --> 00:04:05,000

Geri:
43:39


246

00:04:05,000 --> 00:04:06,000

Yes. So what does management look like practical y? Or leading? Would you say managing or leading?



247

00:04:06,000 --> 00:04:07,000

-



248

00:04:07,000 --> 00:04:08,000

14 -



249

00:04:08,000 --> 00:04:09,000

Transcribed by https://otter.ai



250

00:04:09,000 --> 00:04:10,000

Jofish:
43:48


251

00:04:10,000 --> 00:04:11,000

So I think I do both. But I think of them as kind of separate things. The managing, there's administration of making sure people know how much they're getting paid and things like that, and checking boxes to approve expenses and stuff like that. That's not interesting, right? Trying to build a strategic narrative that we can al agree on, that we're moving towards, particularly in a research world like this, where we don't have a clear path already, has been very difficult and very hard work. I think we've figured out what that is, I think we have some directions to go towards it to validate that this is the right direction.



252

00:04:11,000 --> 00:04:12,000

And that's where it sort of moves over into the leading from the managing, right?



253

00:04:12,000 --> 00:04:13,000

Geri:
44:40


254

00:04:13,000 --> 00:04:14,000

Yeah. So what was hard about it?



255

00:04:14,000 --> 00:04:15,000

Jofish:
44:43


256

00:04:15,000 --> 00:04:16,000

Coordinating with people who have different epistemological assumptions, particularly people who are computer scientists, and have been trained as computer scientists without the rich understanding and, I would argue, even respect for different assumptions about the way that knowledge is created in the world. The thing that I think is HCI's secret power is we're such a broad church, you can do so much stuff and publish it at an HCI conference. And I think that's remarkable. I don't know of another field where you can do so much stuff with so many different kinds of methods, so many different kinds of topics, with so many different kinds of ways of validating the knowledge that you've created in the world, which makes our job very difficult. Because there's no one thing you can say that applies to everyone, right? You can't say, "Oh, well, you've got to have this many N, right, or your p values need to be this. Or you must have this much data or use that algorithm…", because someone is going to do something completely different. And how you can create something that can stil continue to create valid, interesting work in the world, and yet embrace this eclectic mess. I think it's a really fascinating problem. But it's one that the successful approaches to HCI I think need to embrace, not not try and dunk down, right?



257

00:04:16,000 --> 00:04:17,000

Geri:
46:13


258

00:04:17,000 --> 00:04:18,000

Yeah. So you were sort of doing that in almost a microcosm there at Mozil a, with this open space



259

00:04:18,000 --> 00:04:19,000

[Jofish: Very much so.] in a way and deciding what your version of that would be [Jofish: Very much so.] That connected with everyone in your team from the sounds of it?



260

00:04:19,000 --> 00:04:20,000

Jofish:
46:28


261

00:04:20,000 --> 00:04:21,000

Because of people in my team, but then there's other people who are working on similar topics who might be doing slightly different stuff. So we're al working around speech and language, but how you figure out what their contribution is, and how do you figure out if you're having impact in the ways you want to have impact? And I think that question of how do you measure impact, right? In industry, we talk about DAU a lot. DAU is daily active users, and in many, many fields DAU is proportional to everything else, right. DAU is proportional, in particular to revenue, to impact. If you've got a mil ion users, you have much more impact than if you've got 10,000 users, right?



262

00:04:21,000 --> 00:04:22,000

Geri:
47:20


263

00:04:22,000 --> 00:04:23,000

In principle, but it depends what it is that you're offering, what the potential target population is, I guess.



264

00:04:23,000 --> 00:04:24,000

-



265

00:04:24,000 --> 00:04:25,000

15 -



266

00:04:25,000 --> 00:04:26,000

Transcribed by https://otter.ai



267

00:04:26,000 --> 00:04:27,000

Jofish:
47:28


268

00:04:27,000 --> 00:04:28,000

Exactly. I mean, the Rust stuff we're talking about is a nice example. Rust wil never have a mil ion DAU. There wil not be a mil ion daily active users of a systems level computer language. So it's not the only one. And I'm always suspicious of any metric. "Oh, wel , we can measure this. So let's measure this." But I think those questions of how you can have impact, how you can plan to have impact, and how you can measure the impact that you've had. Those are the tricky ones, right? The research grant program, how do I measure that impact? Well, I think it's been great. And I'm working right now with Miriam Avery, our Director of foresight, super smart, comes from real y strong Anthro background, working with her to try and figure out how do we tell a story about this is the value that this has made.



269

00:04:28,000 --> 00:04:29,000

And some of that is direct changes in code that Mozil a ships that came out of this. Some of it's much more theoretical stuff. Some of it might be changes in attitudes in the world. Al of those count, and all those can be real y valid.



270

00:04:29,000 --> 00:04:30,000

Geri:
48:38


271

00:04:30,000 --> 00:04:31,000

And it's up to you to define them and make those arguments.



272

00:04:31,000 --> 00:04:32,000

Jofish:
48:41


273

00:04:32,000 --> 00:04:33,000

Yes, I just put a meeting on my schedule for Friday to have that meeting with Miriam to sit down and say, Okay, how are we going to tell the story of this success? What do we consider success?



274

00:04:33,000 --> 00:04:34,000

Geri:
48:52


275

00:04:34,000 --> 00:04:35,000

So you don't have any Mozil a-imposed metrics of success?



276

00:04:35,000 --> 00:04:36,000

Jofish:
48:56


277

00:04:36,000 --> 00:04:37,000

So there are metrics for the company as a whole, right? We have a daily active user target. We have some other ones around other ways that people use things and engage with stuff and we're counting that. What's the third one, we've just changed some of the details. And so I forget, but there's sort of these very macro level things.



278

00:04:37,000 --> 00:04:38,000

Geri:
48:57


279

00:04:38,000 --> 00:04:39,000

And then for specific projects, it's up to you to real y be clear about what exactly you're contributing to that.



280

00:04:39,000 --> 00:04:40,000

Jofish:
49:32


281

00:04:40,000 --> 00:04:41,000

Real y, in the emerging technologies world, what we're looking for is impact of appropriate thing for the thing we're looking at. Right? So it may be that you have, if you've got 10 users who are real y passionate and using all the time, that's a great sign and I would take 10 users passionately using your product, over 1000 users who like do the thing then they go away, right? That's really that crucial thing.



282

00:04:41,000 --> 00:04:42,000

So trying to find what those signals are. And we know what they look like when you scale up, because you've got nice curves that sort of go up into the right. "Oh, that must be good, the curves going up to the right." If you can think about how do you spot those early on? How do you make the changes that result in that early on? It's real y interesting questions.



283

00:04:42,000 --> 00:04:43,000

-



284

00:04:43,000 --> 00:04:44,000

16 -



285

00:04:44,000 --> 00:04:45,000

Transcribed by https://otter.ai



286

00:04:45,000 --> 00:04:46,000

Geri:
50:24


287

00:04:46,000 --> 00:04:47,000

To keep feeding back into the work that you're doing. Just going back to the thing about how you get the team on board, or how the team col ectively decides this sort of, you know, what they're about, what that vision and story is. What are some of the practical ways that you did that? Did you do away days, did you…? I mean, what did it practically look like?



288

00:04:47,000 --> 00:04:48,000

Jofish:
50:46


289

00:04:48,000 --> 00:04:49,000

So we're a pretty distributed team [Geri: Yes, because Seattle is not just around the corner.] Exactly.



290

00:04:49,000 --> 00:04:50,000

So conveniently, we're al in one time zone now. We weren't for a while. But now we al are, which is a total bonus. Both Tamra and Janice are up in Seattle, Abe is based in San Francisco and comes down to Mountain View a day or two a week, maybe three, sometimes. I go up to San Francisco, I want to say it's every two weeks. And it's not even close to every two weeks. But every now and then. Mozil a as a company does, twice a year, we have all hands, where you have everyone in the company come to one location. Something like 38% of Mozil a employees are not in an office, they work wherever they want to work, but not in a Mozil a office space. Interestingly, when we do the engagement survey, those employees report that they are happier and more satisfied at working at the company than people are actually in an office, which …



291

00:04:50,000 --> 00:04:51,000

Geri:
51:45


292

00:04:51,000 --> 00:04:52,000

Wow, that's interesting.



293

00:04:52,000 --> 00:04:53,000

Jofish:
51:47


294

00:04:53,000 --> 00:04:54,000

But because of that we do get together twice a year there. And we usual y do two other work weeks on sort of the alternate quarters where we'l get together for a week. Maybe it's in Mountain View, maybe it's in San Francisco, maybe it's somewhere else.



295

00:04:54,000 --> 00:04:55,000

Geri:
52:02


296

00:04:55,000 --> 00:04:56,000

Just your team or the whole company?



297

00:04:56,000 --> 00:04:57,000

Jofish:
52:04


298

00:04:57,000 --> 00:04:58,000

This case, is just the team. And we'l try and do short things, like every so often we'l do a week where perhaps three of us or whoever it is, we're doing user testing, we might try and all be in one place for that user testing. Those sorts of opportunities. Those are the in person things. There's a lot of use of Slack, there's a lot of use of, you know --



299

00:04:58,000 --> 00:04:59,000

Geri:
52:28


300

00:04:59,000 --> 00:05:00,000

I was gonna ask about how do you support the communication and the cohesion, and the culture of the team when you aren't face to face.



301

00:05:00,000 --> 00:05:01,000

Jofish:
52:37


302

00:05:01,000 --> 00:05:02,000

Slack and email. The other thing I think is real y interesting. And this one wil be fascinating to come back to in a couple of years and see if I stil believe we're doing the right thing. When we started, there's a term of "Having a Stand-Up", right? In the Agile framing, the stand-up is sort of a 15 minute thing that's literal y standing up. 15 minutes is long, right? It's reporting quickly, say what you're doing, move



303

00:05:02,000 --> 00:05:03,000

-



304

00:05:03,000 --> 00:05:04,000

17 -



305

00:05:04,000 --> 00:05:05,000

Transcribed by https://otter.ai



306

00:05:05,000 --> 00:05:06,000

on. We originally were talking about team stand-ups, and then we changed the names of those.



307

00:05:06,000 --> 00:05:07,000

Because we realized the important thing partly was reporting on the status. But the original notion of a stand up presupposes geographic colocation. Right. And so the standup is this functional thing. But a lot of the soft stuff happens, watercooler conversations happen literally around the watercooler. Well, we don't have that. And so we need to make sure we have those moments. A lot of us have kids, that does make a big difference. Al of our meeting times are framed around -- we are al dropping off kids in the morning and picking them up later. And we work around those times. So having those daily meetings, I think they're on the schedule, usual y for half an hour. Sometimes they go over, I think.



308

00:05:07,000 --> 00:05:08,000

Geri:
54:08


309

00:05:08,000 --> 00:05:09,000

You do them on video?



310

00:05:09,000 --> 00:05:10,000

Jofish:
54:10


311

00:05:10,000 --> 00:05:11,000

On video. And often those wil be, Abe often cal s in from the car because he has slightly later dropping kids off times. So he'l do them from the car on his phone. But it's a nice thing.



312

00:05:11,000 --> 00:05:12,000

Geri:
54:28


313

00:05:12,000 --> 00:05:13,000

That "touch base" every day, hear each other's voice.



314

00:05:13,000 --> 00:05:14,000

Jofish:
54:31


315

00:05:14,000 --> 00:05:15,000

Yes. And there is presumably some sort of optimal amount for this. I wil say al the strategic planning has taken its own toll on us. It's taken a lot of time. It's taken a lot of effort. It's been fun, but it's been a lot of time and effort.



316

00:05:15,000 --> 00:05:16,000

Geri:
54:48


317

00:05:16,000 --> 00:05:17,000

I think Anna Cox, when I talked to her, she had some students who were distributed and she would have sort of almost the equivalent of a stand up on Slack, where they had a bot that would ask them all to put in their morning report about what they were planning to do or what they did yesterday. And at least when I spoke to her, she was saying that worked well, but it misses that real time discussion.



318

00:05:17,000 --> 00:05:18,000

Jofish:
55:11


319

00:05:18,000 --> 00:05:19,000

Yeah. I think it helps once you know what you're doing, it does work very wel . When you're stil figuring out what the narrative is -- I sort of looked forward to our meetings getting a bit shorter. They've been getting a bit long because they needed to be, because we were doing all of this strategic thinking about what we're going to do for next three years. Now, we can just do it.



320

00:05:19,000 --> 00:05:20,000

Geri:
55:34


321

00:05:20,000 --> 00:05:21,000

So it's recognizing phases and adapting process and phases. You've been overwhelmingly positive and enthusiastic about working in industry as a researcher, and doesn't seem like you regret that move at all. Are there any downsides?



322

00:05:21,000 --> 00:05:22,000

Jofish:
55:58


323

00:05:22,000 --> 00:05:23,000

So I think my experiences at Mozil a are particularly positive in that respect. Companies are hard objects, right? There's so many people trying to go in roughly one coherent direction, it's very difficult to



324

00:05:23,000 --> 00:05:24,000

-



325

00:05:24,000 --> 00:05:25,000

18 -



326

00:05:25,000 --> 00:05:26,000

Transcribed by https://otter.ai



327

00:05:26,000 --> 00:05:27,000

steer a ship of that many people in one direction in a coherent way. As I've been promoted, and been in a situation in which I see more and more how that sausage is made. I don't know what the right tools are to do this, to create those narratives in a way that's scalable and has impact across the company.



328

00:05:27,000 --> 00:05:28,000

Geri:
56:44


329

00:05:28,000 --> 00:05:29,000

And that necessarily have to be adaptive for a company to survive in a rapidly changing environment.



330

00:05:29,000 --> 00:05:30,000

Jofish:
56:52


331

00:05:30,000 --> 00:05:31,000

Absolutely. Right. And so I don't know what those right tools are. The tools I feel like I'm using right now are not optimal. And I would like to figure out what those are. And that's been a tiring process. I think I've worked harder in the last two months than I have in the rest of the time I've been at Mozil a.



332

00:05:31,000 --> 00:05:32,000

Figuring out that three year strategic narrative has been real y difficult. And we've worked with sort of various outside agencies, whatever. That stuff's really tricky. I think there are issues of diversity and inclusion, in particular, which continue to be problems across the industry. I think Mozilla has done a better job than most companies around that. But that's only if you count the employees. We have a lot of volunteers who who feed into Mozilla, who are a huge part of what we do. And that's much trickier, right? If I'm trying to improve the diversity of my company, let's say I don't have enough women, right, as a statistical representation of the world. Well, if I work for a company, it's relatively easy in that I can hire more women. That's how it works, right? I don't even have to fire the men, I can just hire more women. And then over time, I wil have more women in the company. If you're thinking about contributions to open source, how do you do that? And we've seen some cases, some pretty high profile cases in which prominent open source projects have realized they have a problem with this, right? How do you make an open source project which real y reflects the community and doesn't exclude people? And I actually think the exciting thing here is to look at things like the work of people like Casey Fiesler, who have been looking at fanfiction communities, which are arguably open communities that require a certain amount of investment to get into it. And you look at the onboarding processes, how do they bring people into their communities, right? Let's see if we can learn from those because they've managed to do diversity in a way that we necessarily haven't. So I think that's a real issue. And something that we need to be thinking about.



333

00:05:32,000 --> 00:05:33,000

Geri:
58:55


334

00:05:33,000 --> 00:05:34,000

That also just sounds like a good chal enge to have. [Jofish: It's a super chal enge to have.] Again, an opportunity to make a difference, have an impact?



335

00:05:34,000 --> 00:05:35,000

Jofish:
59:04


336

00:05:35,000 --> 00:05:36,000

Yes. And I've put some non trivial amount of my time in the last couple of years into that. And it's been satisfying. We had an interesting case where, again, with this engagement survey, people reporting they were being harassed and bul ied at work. And I was like, what? That's because it's real y didn't seem the right thing for the culture, right? So we dig in a little bit more to these anonymous survey results. So we don't real y, we know it's people within ET [?]. And we have a bit more categories than that, but not that much more. It turns out what was happening was that people were getting harassed on the internet, because of the work they were doing as part of their job. So we have really strong technical women who are giving talks as part of our developer relations work. So someone would come



337

00:05:36,000 --> 00:05:37,000

-



338

00:05:37,000 --> 00:05:38,000

19 -



339

00:05:38,000 --> 00:05:39,000

Transcribed by https://otter.ai



340

00:05:39,000 --> 00:05:40,000

and give a talk like CSS grid or something like that right? And then they wil get horrible stuff back like old stupid "Shut up and go back in the kitchen". Right? Exactly. That was my reaction, too!



341

00:05:40,000 --> 00:05:41,000

Geri:
1:00:04


342

00:05:41,000 --> 00:05:42,000

I'm sorry, my mouth, my jaw just dropped. Wow.



343

00:05:42,000 --> 00:05:43,000

Jofish:
1:00:08


344

00:05:43,000 --> 00:05:44,000

And so we're like, "Okay, this is real y an issue". So we spent some time talking to people, we found a couple of services that we felt could help us with this. There's a new company called Tall Poppy, who were just founded by two real y awesome people who have a lot of experience with these kinds of things to help around harassment. So we're working with them. We've got some work with a company called Abine, who have a product called "DeleteMe", which removes your personal information from databases. So let's say I don't like something you've done, I can't look up your address, and then send you 222 pizzas, right? Or a SWAT team or whatever? Yeah, well exactly. So looking at those services to have ways to support people seems real y important.



345

00:05:44,000 --> 00:05:45,000

Geri:
1:00:59


346

00:05:45,000 --> 00:05:46,000

So it's more about supporting your own employees.



347

00:05:46,000 --> 00:05:47,000

Jofish:
1:01:01


348

00:05:47,000 --> 00:05:48,000

It is also because of the volunteer base nature, we're more and more thinking about how we can extend those kinds of protections. And that means writing those contracts in a way that means that we can extend them to volunteers, right? If some volunteer is doing this, because we have a huge volunteer tech speaker program, which is hugely important. How can we make sure that we can look after those people as well? Because they're Mozil ians as well. Right?



349

00:05:48,000 --> 00:05:49,000

Geri:
1:01:24


350

00:05:49,000 --> 00:05:50,000

Again, very values driven work.



351

00:05:50,000 --> 00:05:51,000

Jofish:
1:01:26


352

00:05:51,000 --> 00:05:52,000

Yes, stunningly so.



353

00:05:52,000 --> 00:05:53,000

Geri:
1:01:29


354

00:05:53,000 --> 00:05:54,000

Bril iant. So just in wrapping up, is there anything that I should have asked you that I didn't or that you'd want to be just talking about?



355

00:05:54,000 --> 00:05:55,000

Jofish:
1:01:38


356

00:05:55,000 --> 00:05:56,000

There's one thing I wanted to mention, which is the difficulty that people have when they move from an academic context into industry.



357

00:05:56,000 --> 00:05:57,000

Geri:
1:01:48


358

00:05:57,000 --> 00:05:58,000

Is this straight from PhD into industry or having been in a faculty position going into industry?



359

00:05:58,000 --> 00:05:59,000

Jofish:
1:01:55


360

00:05:59,000 --> 00:06:00,000

-



361

00:06:00,000 --> 00:06:01,000

20 -



362

00:06:01,000 --> 00:06:02,000

Transcribed by https://otter.ai



363

00:06:02,000 --> 00:06:03,000

I've seen both. And I've seen it even for, you know, internships or short things. I'll say this as a big blanket statement. And then you can pick on me for a bit, that'd be fun. In academia showing your work is really valued, right? I had one incident in which I had a colleague I was working with, who had recently come from academia. And we were reviewing an area of academic research, I forget what, doesn't really matter. And he identified about 60 papers that were relevant to this question, sort of put them in a folder, and I was like, great. But 60 papers in a folder is not very useful. So then he went and did a literature review, in which he went through and wrote a paragraph about each one of the papers and said, what was going on. And I was like, Yes, this is stil not useful. Eventually, we got to what we were looking for, which is, here are the three things we should do. We need to, I don't know, paint it blue. And here's the paper that says we should paint it blue. We need to put flowers on it. And here are the three papers that say we should put flowers on it. And we should definitely not call it Dave. And here's the paper that says that, right? And that idea that what was necessary is, what's the impact?



364

00:06:03,000 --> 00:06:04,000

What should we change? What should we do differently? Rather than, How did you get there? Right?



365

00:06:04,000 --> 00:06:05,000

Geri:
1:03:19


366

00:06:05,000 --> 00:06:06,000

The actionable stuff? Give me the sort of "So whats". [Jofish: Yes.] And some evidence that you've got the stuff to back it up. So yes, I can total y relate to that. Because I used to work in industry, when I think back to some of the deliverables that we'd produce for clients. They were rigorous and detailed.



367

00:06:06,000 --> 00:06:07,000

And even though we worked with a graphic designer, who was very good at sort of listening to us explain what we thought was important and putting it into a very simple graphical model, we stil packed it with too much text. We did try to do the executive summary that tried to pull out the things but even they were quite dense. And I would just approach it so much differently now. So yeah, how do we get that transition to actual y shifting our understanding of how we, again, communicate results to have impact?



368

00:06:07,000 --> 00:06:08,000

Geri:
1:04:11


369

00:06:08,000 --> 00:06:09,000

And in fact, that's what we did end up doing more as we learned over time, was actual y doing the slide deck. I don't think we were so good in putting a lot of the backup data in the back of the slide deck. And I can see that that would work real y wel , because it also means that someone who's in charge of that division, who cares about the recommendation on slide three, can go into the detail that's related to slide three, if you can find some way of making that path clear to them.



370

00:06:09,000 --> 00:06:10,000

Jofish:
1:04:11


371

00:06:10,000 --> 00:06:11,000

I wil give one explicit thing that I think is valuable. I can go and find a dozen recommendations on how you should give talks, and what should be on slides, right? And they're all like, keep all the text off them, keep them in a more, whatever. What that ignores is the reality that in the corporation, the place where knowledge resides is in the deck, right? It's in the Google Slides. It's in the PowerPoint slides.



372

00:06:11,000 --> 00:06:12,000

What that means, however, is the right way to do it is not to put al the data on the slides. It's to separate out the archival and knowledge repository part of the deck from the bit that you present.



373

00:06:12,000 --> 00:06:13,000

Right? And so you explicitly say, here's the 10 slides that I'm presenting to you, and maybe within that you link to things. But here's the 10 presentation slides that happen to be in the same slide deck as, here's the tables of data, here's the list of questions. And those can be approximately speaking as much data as you want to put on there same as with any other thing. And I'd love to say that we could



374

00:06:13,000 --> 00:06:14,000

-



375

00:06:14,000 --> 00:06:15,000

21 -



376

00:06:15,000 --> 00:06:16,000

Transcribed by https://otter.ai



377

00:06:16,000 --> 00:06:17,000

put that data somewhere else in a written document. But I've found that that doesn't work so well. The coin of the realm is the slide deck, right?



378

00:06:17,000 --> 00:06:18,000

Jofish:
1:05:55


379

00:06:18,000 --> 00:06:19,000

You need to have reliable, robust data. I'm not disagreeing with that. You just need to say, by the way, it's reliable and robust. SEE SLIDE 52. In which we talk about how reliable and robust this is, right.



380

00:06:19,000 --> 00:06:20,000

Geri:
1:06:11


381

00:06:20,000 --> 00:06:21,000

Rather than read the next 20 pages? [Jofish: Yes, yes.] I think that there is an increasing move in a lot of academic contexts for funding bodies to require reporting on impact, and they're getting increasingly particular in how that's reported. And in having impact, people are having to go out and speak more to the public or think about those public engagement strategies. So maybe in general, people wil get a little bit better.



382

00:06:21,000 --> 00:06:22,000

Jofish:
1:06:44


383

00:06:22,000 --> 00:06:23,000

I hope so. Then again, I think things like, the medium posts are a great way to have that.



384

00:06:23,000 --> 00:06:24,000

Geri:
1:06:50


385

00:06:24,000 --> 00:06:25,000

Yeah. I'l make sure to put a link to al of those on the webpage. [Jofish: That'd be great.] Yeah. Wel , Jofish, thank you very much for your time. Really appreciate it.



386

00:06:25,000 --> 00:06:26,000

Jofish:
1:06:59


387

00:06:26,000 --> 00:06:27,000

I real y enjoyed it.



388

00:06:27,000 --> 00:06:28,000

Geri:
1:07:05


389

00:06:28,000 --> 00:06:29,000

You can find the summary notes and related links for this podcast on www.changingacademiclife.com.



390

00:06:29,000 --> 00:06:30,000

You can also subscribe to Changing Academic Life on iTunes and now also on Stitcher. And you can fol ow ChangeAcadLife on Twitter. And if something connected with you, please consider sharing this podcast with your colleagues so that we can widen the conversation about how we can do academia differently.



391

00:06:30,000 --> 00:06:31,000

-



392

00:06:31,000 --> 00:06:32,000

22 -



393

00:06:32,000 --> 00:06:33,000