Skype was one of the first multi-platform communication service even before the WhatsApps, WeChats and Vibers of the world became cool. However, one feels that Skype didn’t get the love it deserved inside Microsoft and despite having over 300 million monthly active users, it never got the same attention like Apple’s FaceTime. That is changing now under Nadella’s Microsoft and Gurdeep Singh Pall, who now heads the Skype team is keen to ensure Skype becomes cool again.
I met Pall at his office at Microsoft’s headquarter in Redmond, Washington. Just two days before our meeting, he had unveiled Skype Translator that translates video calls from one language to another in real time. It definitely looks like something you’d expect to see on the sets of Star Trek, not in real life.
“I used to manage all the speech efforts at Microsoft until six months ago when I switched to Skype. We were on this linear path and then suddenly out of the blue this technology called deep neural network which had been shelved away for 20 years, suddenly raised its head again. In the 80’s, artificial intelligence was a big thing and was funded by Office of Naval Research, NSA, and all such big agencies. A lot of work happened in that field and that sort fell flat on its face. The reason was because people had not thought about machine learning, and rather thought about rules based AI,” Pall explains.
Deep neural networks is a framework that shows how machines can learn on their own if you keep providing it with data inputs to consume. The more data that is fed, the better it becomes. The model had existed for almost two decades but was dismissed by most researchers. However, Microsoft Research kept working on it and in 2009 researchers hit a breakthrough that led to incremental speech recognition. This is the underlying technology that is now powering most voice recognition services including Cortana, and now Skype Translator.
But Skype Translator is just one of the things that Pall is excited about. He is now bringing the Lync and Skype efforts under one roof, since both communication services do similar jobs for different audiences. “I’m going to have one offering from the individual to the enterprise under one brand, one set of experiences which works differently and its capabilities are framed in the right way,” he tells me. You can hear the excitement in his voice and passion burning in his eyes.
He promises there’s a lot of stuff that’s happening at Skype and there are many announcements in the pipeline. We discuss what’s next for Skype and competition from the likes of WhatsApp and others. Here are the edited excerpts from the interview.
Let’s talk about your new role here at Lync and Skype. How is it that you see the two products that typically do almost the same thing — messaging and communication collaboration — but have been quite separate?
From my perspective, I have the best job in the universe and here’s why. Look at the industry today, and point out one place where the same tool is used by individuals, group of users and used by enterprises. It is very hard to find one. I will tell you one, it is called the Microsoft Office. In the last 20 years, nobody has been able to recreate this magic, and that is what I’m going to do. I’m going to have one offering from the individual to the enterprise under one brand, one set of experiences which works differently and its capabilities are framed in the right way.
For example, on the enterprise side things like compliance and security and scoping and these features are important, while on the individual side, privacy is important. So imagine like a freemium model on the consumer side and a subscription-based model on the enterprise side all provided under one brand. So after Office, we can do this and that is my mission.
Let’s talk about Skype, it is only now that we seeing all these messaging apps’ valuations going crazy. Everyone thought when Microsoft bought Skype, the valuation was equally crazy back then. Nothing really happened with Skype, in the sense, it did not become the go to messaging or communication service like how WhatsApp has become in many countries and India in particular. People there have stopped using text messages and the very first thing they use is WhatsApp. Skype has been there for long, but it has not become that go to service.
I’ll agree with you and I’ll also disagree with you. I think what you see WhatsApp as doing is disrupting SMS. Skype is what’s disrupting voice and video calls. So they are slightly different. Now you can say they are adjacent, and they are good questions and we will keep working in that space. The great news is that comms itself has become very exciting. Email has, at least in the consumer world, has lost a lot of the action. But the space has grown huge. Skype is a huge contender at this. I mean I see all the numbers, so I think we have a really good platform to more things with it. In my opinion it is harder to go from not rich to rich in comms, it is easier to go the other way.
But what happened before you took over this particular group? Has the focus changed, especially for Skype, considering it was a more consumer-centric product and service. So what has changed or is changing in terms of how you look at Skype as a product?
I was involved with the acquisition of Skype, before I moved over to do this AI work. The vision always was that it was a great consumer asset as well as a very good enterprise asset and that they could be put together in a very meaningful way. So that has continued and with my coming in, I’m only making sure that it is very clear how these things are going to come together. I believe, like Office, this is can be a very enduring product line if we can bring them both together.
A lot of messaging trends that we are seeing is, suddenly your phone number, especially for first time Internet users, is becoming your id. I believe for the new Microsoft, the theme is to capture these first time Internet users right away. A lot of them don’t have an email id or an online presence, but all of them have a phone number. So are you looking at phones becoming a sort of first gateway to the Internet and how is Skype changing to take advantage of that?
We are taking that very seriously.
Can you share what you are planning on doing?
It really comes down to the notion that, is the phone identity important. Our belief is that it is a good on ramp identity, and it breaks down over a long period of time. I’ll give you an example. I use too many phones, and I use all these apps and trying to get the apps across these many phones across these many screens is really a pain. The fact that I can use my email identity across all these phones and have all my comms and history there is really useful. We are headed towards a world where I think there is a lot of projections, actually Mary Meeker’s new projections are quite good, and we are looking at five devices per user, so that is 10 billion devices. So the moment you go to five devices per user, the phone number starts to break down very fast, so I think my personal view is that this isn’t an enduring trend. However, it is an important trend that we cannot ignore.
Talking about translator, it is still pre-beta, so what is the roadmap? Which languages are you looking at?
We have lots of languages. If you look at it, there’s actually no magic in it. If you look at the languages we support for Xbox One or Cortana and if you see our speech APIs and if you see the languages we do machine translations in, you will get a fair idea about the languages which are all candidates. So now the question is when do we roll out which language. There is a quality bar we want to hit and this is true for all machine learning things.
I don’t care if it is for search or image recognition, machine learning has the same formula and it looks like this. There is an improvement based on the amount of data you have and you don’t want to roll out a functionality where you are not good enough. But a lot of improvements actually comes after you hit the bar and lots of people start using it. So the right time to release it at scale is a time when you hit the main bar and you then fan it out. So with this criteria we will apply the same formula objectively for all the language sets we have and the moment we hit that we will go.
The good news is we have lots of data with an opt-in sort of way and we do more than 2 billion minutes of calling per day on Skype.
How do you see the smartphone uptake, especially in what we would call the underdeveloped countries, communication is typically more important for them with someone in the family earning his living outside the country?
It is amazing how Skype is used in country say Philippines. I can tell you which hour or day of the week, Filipino workers are communicating with their families. I can tell you because we see all the regional patterns of comps and things like that, so your questions are how do I see the penetration of smartphones.
Yes and considering that Skype is not there in feature phones I don’t see a $20 unsubsidized phone doing Skype.
I would say Moore’s law is our friend. I mean it is actually crazy if you could look at some of these low-end smartphones. The prices are extremely low, and there are just getting lower. I mean, you can just do the math, it’s going to be half the price in about 18 months. I think that part of the equation we feel pretty good about, the piece which is bit of a wild card is the bandwidth and the connectivity costs. We are seeing hyper competitiveness in India and countries like that where the costs are aggressively going down, so if that continues then it will be fine but I do worry about that piece little bit.
I remember you had a Skype Phone with 3 in the UK. A lot of messing service are tying up with carriers, Facebook and WhatsApp tie up with carriers, where they offer a Rs 30 monthly pack that gives you access to Facebook or Twitter. So is that one of the ways while not looking at rich comms but at least on messaging bit?
We look at all these ideas and you know we discuss this a lot. A lot of this depends upon carrier too and what we want to do, and are the carriers open to it. I think it’s a very shifting space for them as well. Forget Skype for a moment, look at what WhatsApp has done to the SMS revenue. They are sort of trying to figure out their world and trying to make sense of it, we watch that and we look at whatever things we want to do.
But is there something that is probably a light version of Skype for feature phones, or phones that can connect not giving them the rich video quality?
We are actively thinking about those things.
A year down, what is it that you envision Skype to be, how different would it be from what it is right now?
There are few dimensions — one is that we have talked about all the way from work and life, that is going to be the big thrust for us to move forward. We are doing really well on mobile, last year we have really started doing really well. At some point we will share our numbers on Androids and iOS and they are actually very impressive. So that is going to be a significant step.
The other thing you will see is really look at mobile at much more richer way, in the sense that all the things we are talking about are part of a sort of multi-dimensional equation that need to be solved. It includes geo, bandwidth, what does the app look like, and is it broken into pieces and that is essentially what we are doing now. It is a bit too early to share it with you now, but expect us to do bunch of things in that phase.
You have iMessage on iOS, which is the default to message if the other person too is on an iOS device. Android gives you Hangouts now. Skype, on the other hand, is in very good position considering it is available across platforms. Additionally, you also now have Windows Phone, Windows 8. Why doesn’t Skype become the official messaging app for Windows? Is that something you are looking at?
Firstly, Windows 8.1 and Windows Phone 8.1 has Skype in it.
But it’s not the default messaging app.
Yes it’s a separate app. You can expect to see a lot more integration.
How is going with the other platforms? Are they being developed at the same pace or do we see features coming first to Windows and Windows Phones and then going to iOS and Android?
We are very committed to cross platform, the work happening on Android, iOS, etc. is really we are doing some amazing work. We are just going to be working on multiple fronts and our commitment is to all the platforms and we will do a great job on all platforms. If you look at iOS, it is such a close system and you really have to do so much extra work to even get the basic experiences right and till today iPhone locks out other apps from accessing the hardware for doing acceleration. I think it is such a short sighted strategy, but that’s where they are and the market is starting to teach them lessons that that is not the model.
I think Android is better but Android is chaotic with you know 54 versions of Android but if we do our best to identify the major ones and do a great job there. On Windows we do great as we have access to pretty much the whole OS, so I think you should expect some things from us there. On the other ones, we will do the best we can as much as the platform allows us to.
Now that you are moving to the London HQ, how does that help you move things forward faster?
Skype is pretty distributed. I have a lot of people here, Lync is here, so I’m going to practically spend half my time there and the other half here.
Where are all your development centres? Is there anything happening out of India as well?
There’s nothing in India at the moment. I have had teams in India in my previous gigs, but nothing at the moment. My dev teams are in Tallin, Stockholm, Prague, London, Paulo Alto, Redmond, Moscow and a small team in China as well.
Is there any trend in the messaging space that is really exciting you right now, which Skype doesn’t have now?
The most interesting thing that I see in comms has to do with video and time shifting in videos. The reason I say that is because every now and then people will take a step back in functionality, because what they get is a short term optimization as something else opens up that is interesting. So text messaging for example, for it is something like going back to when I could do something more convenient because I’m trying to do somewhat of an arbitrage between various things including costs, time, etc. That comes quickly, but what after that? You never forget the fact that you could have a conversation with this person in 1080p video, but now that I have got this, what is the next adjacent move that I need to make from there?
When you think like that, with rich video, time shift like experiences will become very important. All the research we have done show people connect better when they hear the other person and event better when they can see him/her. In my opinion that is the shift that you will see and we will get out of this frenzy of hey I was able to do text messaging for free. It is more of a passing fad, and you will see a lot of new and interesting stuff.
I met WhatsApp’s Jan Koum in February and his idea was that he wants WhatsApp to be able to do just one thing — that is clutter-free text messaging. Over here we are talking about rich comms and also giving the option that if you want the uncluttered, then you can always just text someone. So do you see that they are two different schools of thought, or how would consumers see that?
No one has seen the future, and there is no point in seeing the past and predicting the future. Nikon for me is the king of photography, but today who do you think is more important in photography — a smartphone or Nikon?
If you are talking about the masses, then it is the smartphone.
So basically you end up at a place, if you do one and only one thing very well, you could end up with a very small usage share. Jan’s a smart person and he has done a very good job, but this is a space where you don’t want to be very convicted about anything. You just want to be able to keep changing course as more things become apparent and I have to say we too have had a share of our own issues.
Skype has been around for so long and it has been central to most of messaging that went out with Windows initially, when Windows Phone came out even then. But there has not been much buzz about Skype until earlier this week.
You can’t pay for that kind of buzz like when we made group calling free and Skype Translator. All I will say is watch this space. You should expect a lot. Have you seen Skype on Xbox One? It is by far the most amazing experience. We just shipped a snap view, wherein you can literally have five people on a call with you watching the same show and having a conversation. You can be watching a show with a video of your friend snapped to side of the screen, watching his/her reactions. To me there are a bunch of things and you are going to see a lot from us.