When I sat down to interview Jane Silber, CEO of Canonical, I don’t think it was lost on either of us that our ability to chat freely even though I was in my office in the middle of the U.S. and she was in her office in London, England had everything to do with cloud computing, an area in which her company does brisk business.
Silber has been running Canonical (maker of Ubuntu, among a great many other software products) in one form or another for well over a decade at this point, first as COO and now CEO. She answers questions thoughtfully, with carefully chosen words; even though I’m sure I’m not the first journalist to ask her some of the below questions (maybe not even the first one this week), she had no canned responses, and she never veered off course to discuss her own agenda. There were no preset talking points; simply, I asked questions, and she answered them.
The impression I got from speaking to Silber is that she has a broad range of expertise--she seamlessly pivoted from discussing management styles to explaining the specifics of Canonical-made technology--and doesn’t like to be too far away from what’s happening on the ground. (I noticed, for example, that she was hanging out at the Canonical booth at MWC 2016. Most CEOs don’t spend much time on the show floor at those sorts of events.)
From across an ocean, we discussed her background and career path, the emergence of Canonical, the changing tech landscape, and girls and women in STEM fields. (Cameos: artificial intelligence and virtual reality.)
Tom’s Hardware: Your resume is an interesting mix of business and tech and science. What was your first interest among those fields, and how did you come to it?
Jane Silber: Originally, the tech part--I got to that through math. I don’t think I took a computer class per se until college. But I’ve always been interested in that math/computer science/engineering mindset. I like puzzles, I like logical thinking, and that’s what pulled me there. And then I discovered programming and the business lens of computer engineering.
TH: So for computer engineering, you were more on the hardware side than the software side at first?
JS: Not career-wise. Certainly in my early college-level education I was interested in that. From a career perspective once I was out of college, it has been software all the way. My first job out of college was with a startup, and I was doing software development. It was the kind of a startup that had written their core product in a language that the CTO of the company had written as a PhD dissertation. So nobody one in the world knew it. [laughs] By the time I came along, they realized that maybe that wasn’t the best idea. So [I had to] learn that language and translate it into something other people knew.
Then I had many years as a developer. But it was the tech and software engineering that caught my interest. And as I progressed in my career, I became more interested in the relationship of tech to business, and the other disciplines within the company and how all the people fit together.
TH: Tell us about how you went through your education, your choices for graduate work, and how that led to your presence role at Canonical.
JS: My undergraduate degree was math and computer science--not a double major, but a joint degree of math and computer science. I worked for a couple years at the startup that I mentioned. I had worked there as a summer job while in university and then went back again for a couple years afterward.
After a couple years of doing that, I wanted to really explore more about the relationship of technology with other aspects of running a business, and frankly, society in general. So I chose a graduate program that was called “Management of Technology,” but it wasn’t a management course per se--it was more [about] how technology was integrated into society and business. I did a special concentration in AI (artificial intelligence), which I found fascinating both from a technical perspective and from that intersection of society and business, as well.
While I was there, I was working as a research assistant on a project that was funded by a Japanese company. When they found out I was graduating, they offered me a job in Japan. And I’d never thought about working in Japan before, but...I thought I’d be silly to turn down that opportunity.
And so I moved to Japan and had that job in a special R&D lab. They set up this somewhat elite R&D group to do product development and research and figure out how they [could] develop a software line of business. I specifically focused on looking across the range of their product ideas, and seeing where I could apply artificial intelligence techniques into those projects and help differentiate them.
It was a great experience at all sorts of levels--it was a great international experience, it was my first management role.
After about two and half or three years, I came back to the US., and to be quite honest, I just needed a job! [It wasn’t a] planned career progression. I took a job as a developer, thinking that I’d do it for six months while I figured out what I wanted to do. It was a small defense contractor that I joined. And it ended up being really interesting, from a technical perspective, from a growth perspective. That was the role that I really grew the most in. I ended up being there 8 years.
I was Vice President by title. The founder was the CEO, but in many respects I was running the company, and that was my “stretch” job. That was the one where I kept being offered challenges of things I’d never done before and had that moment of, “Geez, I’ve never done this before, I don’t think I can do it.” And then you do it. You figure it out.
It was a great experience. We sold that company to General Dynamics [GD], and then I spent a couple of years running a business unit within GD. And then I decided I needed a change again. I liked academics, I liked the learning environment. So I decided to go back to graduate school again, and this time I did an MBA and I came over here to England and was at Oxford. Following that, Mark Shuttleworth and I ended up at Canonical.
TH: What was your first job/role at Canonical?
JS: Title-wise, I was COO. Mark Shuttleworth was CEO. I wasn’t there at the very beginning. He spent the first couple months recruiting Debian developers, so he already had about 15 developers (virtually) gathered by the time I joined. My first role, my first set of responsibilities, was to help turn it into a company. [laughs] Make sure people got paid, figure out what we were doing. And in those early days, I was sort of the marketing department--and the sales department and the admin department and the HR department and the finance department, and all those things you do in a startup, wearing lots of hats.
TH: I think it’s really interesting when someone like you, in management roles for many years, starting from scratch with a small company like that. It seems scary...what’s your thinking when you take on that challenge?
I think that’s the most exciting part. It’s a chance to create something. People that write software have this drive to create something, to solve a problem. There’s that same challenge to me around creating a well-functioning company and having it grow, and solving a business problem, and finding a market that you can address in some way.
I mentioned that I like puzzles. In many ways, it’s a multidimensional puzzle where all the puzzles have to fit together. And sometimes, some of those pieces are working really well together, and sometimes the wheels aren’t turning as well.
Those early days [of Canonical] were exciting! I definitely like a smaller company environment as opposed to a larger company environment; the speed with which you can operate in a small company, the impact that you can have, I think is much greater. There’s much more room for self direction and autonomy (in general).
TH: What areas are driving Canonical’s growth?
JS: The largest areas are around cloud, and specifically, a lot of engineering capabilities in product development, in sales engineers, in support teams--almost any technical discipline that you can think of. And that’s driven by customer growth and business growth around Open Stack offerings and the use of Ubuntu in the public cloud. And Juju--which is probably the topic for an entirely different [conversation].
“Little Servers”
TH: The tech landscape has changed dramatically in the 12 years since you started at Canonical. What are the major changes you’ve seen, and how has Canonical been able to address those?
JS: I think from a broad perspective, the rise of cloud [computing] and the impact that’s had both on the architecture within enterprise IT organizations and the economics that people expect: the ability to pay by consumption rather than prepay for general capacity. I think that’s been the real shift that wasn’t present in the early days.
The other big trend these days is IoT and smart connected devices--almost as a carry-on from mobile. When I look back at where we started with Ubuntu as a desktop operating system (and then we added in the server around 2006), versus the type of opportunities and issues that we’re facing today around the cloud and connected devices, and what we think of as the re-emergence of a converged, personal computing experience, it’s really a very different world.
Even in terms of the mindset around the technology and open source in general has shifted dramatically around this time period. I remember when we started, we would start many customer meetings explaining what open source was and talking about licenses, and allaying fears of customers around open source--and none of that happens anymore. There is such broad acceptance that open source is not just the credible way, but a better way of writing and procuring software.
TH: What do you think about the current state of Linux in the current tech landscape, and where do you think Canonical fits in?
JS: I think part of that is that wide acceptance of open source: Linux is everywhere. And I’m sure you’ve seen all the stats from groups like the Linux Foundation and the broad range of services and products which it powers. And I think Ubuntu is a key part of that--certainly in the cloud, Ubuntu is the dominant Linux operating system. Most of the scale-out services that you’re familiar with are built with Ubuntu. Everything from Netflix, to Spotify, to Instagram--all of those are built on Ubuntu.
There are certainly other players there, but in today’s world of scale-out computing in the era of this software, I think Ubuntu is the dominant choice. That’s something we have courted and worked towards, from both a technical perspective and a business model perspective--seeing that the world is moving to doing everything to scale, that the interesting work there is these complex collaborations of software and scaling capacity.
The other side of the coin is when you get down to devices and you look at personal computing and the emerging world of IoT. In the world of IoT, it’s interesting to think whether these smart connected devices are more like phones or more like little servers. And I think in our view they’re more like little servers.
We’re not trying to address the tiny embedded world. We’re not expecting to see Ubuntu on [smart] light bulbs, for example. But the control panel, and the controllers for these light bulbs that would be connected to the internet in some way, I think is a ripe and growing field that really looks a lot like little servers. And that’s an area that we know how to address really well, and are excited about the innovation that’s happening there.
On Ubuntu Core And Convergence
TH: I want to make sure we talk about Ubuntu Core. What I saw at Mobile World Congress [2016] really impressed me. Microsoft and Apple and Google are tackling this idea of convergence, and there are pros and cons to all their solutions. How is Canonical’s approach better or worse?
JS: It’s still relatively early days in this area. One of the things that we’ve done is to consider the unified platform and convergence from the ground up. So, we aren’t just taking a “make the UX the same across [platforms]” approach. We are looking at the fundamentals of both the stack and the operating system, and driving a converged user experience such that you get a similar, but not necessarily identical, user experience on different devices.
We’ve taken some key architectural decisions early on around security and the security models that I think differentiates us that applies to app isolation, that applies to transactional updates and rollbacks of the key parts of the operating system and applications on top.
I think the UX concept that we’ve designed is also very good. We still have work to do to bring that all together, but from the testing user research that we do, we’re very excited about that. I’m not as familiar with the developer tools that some of the other platforms are offering, but the developer experience, across the range of platforms on Ubuntu, is really good and will get better.
TH: It sounds like you guys still need some developer buy-in, but it also seems like the requirement from developers [to get their app working across platforms] is fairly minimal. Is that accurate?
JS: Yeah.
TH: For them, it’s mostly just UI tweaks?
JS: Right. For the phone/tablet space, we have an SDK, which addresses UI elements of having an application sort of re-flow to a user interface depending on whether you’re on the phone-size screen or the tablet-size screen. So it’s up to them to make the decision on how they want that to happen, but that’s the common assumption--that the SDK provides a lot of help.
We also are doing some more in the Ubuntu Core world itself, around the Snappy architecture--Snaps being the new packaging format, to allow easier packaging and transactional updates of applications, independent of the operating system, which is something that has been a struggle historically in the desktop world, and that’s a real boon for developers. It allows developers and ISVs to be in control of their own updates and release schedules in a way that wasn’t previously possible in the classic Ubuntu desktop.
TH: Some people are confused with the nomenclature around Ubuntu Core and Snappy; can you clarify it for us?
JS: So, Ubuntu Core is the minimal edition of Ubuntu, which is largely targeted at smart connected devices. It’s based on some new architectural constructs which have been nicknamed “Snappy.” Sometimes we refer to Ubuntu Core in a nickname way, where “Snappy” is an adjective describing that architecture.
It is in contrast to what we call the “classic” Ubuntu architecture, so it is the Snappy architecture that allows us to do transactional updates and image-based updates and rollbacks and things like that. [It also] provides greater security through app isolation and other architectural decisions.
So Snappy is an adjective or a nickname that describes that architectural construct rather than a product name, and the product name is Ubuntu Core.
The Future: AI? VR? And Operating Systems
TH: It’s funny that you mentioned AI at the beginning of this conversation. As you’re taking on tech giants with your unified platform, there’s also Siri, Cortana and Google Now to consider. Is Canonical developing its own AI to go along with its unified platform?
JS: We are not working on that at the moment. We recognize [AI] as a place where right now the Ubuntu products don’t have as good of an initiative as some of those competitors. There are some open source projects that provide that sort of capability, [but they] haven’t really felt like they provide a full-enough experience to include by default right now. It’s not something that--right now--we intend to take on as a new development project.
I think it will be interesting to see how the standards and expectations of user experiences evolve. I think it might be more likely that we enable for virtual reality and upcoming methods of interacting with the computer, rather than trying to play catch-up on voice recognition.
TH: You mentioned VR. What sort of applications do you think Canonical might be able to work with in that medium?
JS: The way we would likely approach that is to open the capability for developers, and we’ve got the preliminary work in our display server (Mir), to understand what the implications are there. We have some conversations ongoing with partners to explore that, but it’s pretty early days. There’s not an upcoming product announcement that you’ll be surprised with.
Just yesterday I put on the Vive headset for the first time. It’s really incredible. For us, in our world of convergence, we believe that the personal computing experience doesn’t need to be defined by the size of piece of glass that you’re interacting with. Right now, generally people think about phones and tablets and laptops; that’s their norm of personal computing. But the vision of convergence that we’re working toward, where you basically have the same platform and an adaptive user experience, can extend to not having a piece of glass at all. You could be carrying around your computer in your pocket, and interacting with it in an AR sort of way, rather than swiping your finger across a four-inch or 10-inch or 15-inch piece of glass.
[In VR], I think people don’t want a series of standalone experiences. There’s a continuity that people expect in their personal computing experience. I think it will evolve beyond a bunch of standalone gardens.
TH: So really, you see VR/AR as just another component of these unified platforms, or will be, anyway.
JS: That’s exactly how we see it.
On Women And STEM Fields
TH: Women are typically underrepresented in STEM fields, and also in the readership of publications like ours. What can be done to curb that trend?
JS: Unfortunately, I don’t have the panacea, silver bullet answer. I think it’s a multifaceted problem that needs a multifaceted answer, and I don’t think any of the simplistic answers work.
I think some of it starts before people even hit the workplace. I think there’s certainly a big dropout rate for girls and young women in STEM, in education environments. Some of that comes out of a really unfortunate self-fulfilling prophecy. I talk to girls in high schools sometimes, and they say things like, “I’m not going to do computers in college because it’s full of guys.” It’s really frustrating to hear that sometimes!
I think the answers there are largely cultural. There’s lots of studies about the cultural and societal impacts around there. The advice I usually give girls and women is to focus on what they enjoy and be confident and go for it. I think it just takes...doing it. Figuring out how to be in that environment in a manner that’s authentic to them. I think people sometimes feel pressured to do something that doesn’t feel authentic to them; everybody should be able to pick their own path.
TH: Do you think it’s changed at all from when you were first getting your education and getting into the field?
JS: There’s certainly more awareness and discussion around the issue. There’s even an acknowledgement that it is an issue. When I first started getting into my career, it wasn’t even really discussed.
I did find early on a mailing list of “women in tech” called “Sisters,” and it was from what is now the Anita Borg Institute. It was women in systems at various levels, and I found that to be sort of this rare place where women were talking to each other. I certainly didn’t have many female peers, and if you look at my career, I did tech in Japan, which was very male dominated. I did that startup in the U.S. that [General Dynamics] bought; it was defense contracting, which is a very male-dominated area in terms of the companies and the customer set.
I’ve been pretty fortunate in my career. I’ve certainly been witness to what I would consider sexist behavior, but not of the horrendous, egregious type that the some women suffer from. And I’ve been able to, for myself, find a way to address those issues in a way I felt they needed to be addressed, and in a way that made me comfortable and made me confident to continue on. So I think that’s a tricky balance.
TH: One thing that keeps me awake at night is me, and people like me, [not] being aware of blind spots that we may have, [such as] the language that we might use. What are some of the blind spots that you see people have when it comes to gender and sex in these tech fields?
JS: That’s a very good question. There’s so much just built into society and culture today, there’s so much that’s built into our language that we use. These aren’t necessarily things where I would personally go and call someone out on, but they’re very present, from--what I personally think is somewhat innocuous are words like “chairman” (although I personally will say “chair of the board” rather than “chairman.”)--to phrases that get used in the business and tech world all the time. For example, we’ll [say] whether someone has the “balls” to go after something. I’m trying to think of more examples. [laughs] Telling someone to “man up”--meant to convey, “Come on, you just have to get this done, it’s really important.” But the language used is, “You need to man up. A real man would be able to do this.”
I think people may have a blind spot to just how widespread the messaging is that looks are important, that men are better. It’s very present in many ways. It’s depressing if you dive too deeply into it. [laughs] But it’s certainly there.
Jane Silber is the CEO of Canonical. This interview was edited for length and clarity.
