Microsoft Research Podcast

Microsoft Research Podcast

An ongoing series of conversations bringing you right up to the cutting edge of Microsoft Research.

Game on with Dr. Chris Bishop and Phil Spencer

June 19, 2019 | By Microsoft blog editor

Phil Spencer and Dr. Chris Bishop

Episode 81, June 19, 2019

Dr. Chris Bishop is a Microsoft Technical Fellow and director of MSR Cambridge, where he oversees an impressive portfolio of research including machine learning, AI, healthcare and gaming. Phil Spencer is the Executive Vice President of Gaming at Microsoft where he oversees everything from the design of the next Xbox console to the creation and release of blockbuster properties like Halo, Gears of War and Forza Motorsport. These two powerhouse executives are pushing the boundaries of creativity, technical innovation and fun across the spectrum of gaming genres and devices for nearly 2 billion gamers around the world.

On today’s podcast, Chris and Phil discuss their respective roles in Microsoft’s gaming ecosystem, revealing a sort of “enrichment pipeline” that flows all the way from researcher to developer to gamer. They also give us an inside look at the close collaboration between the world-class research organization of MSR and the world-class gaming franchise of Xbox, highlighting Microsoft’s unique ability to deliver the tools, talent and resources that fuel innovation and help shape the future of gaming.

Related:


Transcript

Host: Welcome to the third episode in our recurring series of two-guest podcasts. This episode is all about games. Join us as we follow ideas from the far out, “what if” world of MSR’s gaming research to the far-reaching, “here’s what” world of Microsoft’s Xbox gaming franchise, and see, from the perspectives of both a research executive and a gaming executive, how ideas collide, worlds come together… and everybody has fun.

(music plays)

Host: You’re listening to the Microsoft Research Podcast, a show that brings you closer to the cutting-edge of technology research and the scientists behind it. I’m your host, Gretchen Huizinga.

Host: Dr. Chris Bishop is a Microsoft Technical Fellow and director of MSR Cambridge, where he oversees an impressive portfolio of research, including machine learning, AI, healthcare and gaming. Phil Spencer is the Executive Vice President of Gaming at Microsoft where he oversees everything from the design of the next Xbox console to the creation and release of blockbuster properties like Halo, Gears of War and Forza Motorsport. These two powerhouse executives are pushing the boundaries of creativity, technical innovation and fun across the spectrum of gaming genres and devices for nearly two billion gamers around the world.

On today’s podcast, Chris and Phil discuss their respective roles in Microsoft’s gaming ecosystem, revealing a sort of “enrichment pipeline” that flows all the way from researcher to developer to gamer. They also give us an inside look at the close collaboration between the world-class research organization of MSR, and the world-class gaming franchise of Xbox, highlighting Microsoft’s unique ability to deliver the tools, talent and resources that fuel innovation and help shape the future of gaming. That and much more on this episode of the Microsoft Research Podcast.

(music plays)

Host: So, we have a dynamic duo in the booth today for a special two-guest episode of the Microsoft Research Podcast. I’m sitting here with Chris Bishop, who is a Microsoft Technical Fellow and Director of Microsoft Research in Cambridge, UK. And I’ve got Phil Spencer who’s the Executive Vice President of Gaming at Microsoft and he leads the company’s business across all gaming, devices and services?

Phil Spencer: That’s right, yeah. And I am the duo part. Chris is dynamic. I am duo.

Host: Well, welcome Chris and Phil. Phil and Chris, to the podcast!

Chris Bishop: Great to be here.

Phil Spencer: Yeah, it’s awesome, thanks.

Host: I’d like to start on sort of a philosophical note to give us a framework for our topic today. From the cultural historian, Johan Huizinga, whom I share a last name with, he famously said, “Play is older than culture.” And uber-nanny, Mary Poppins – from your hometown, Chris – famously said, “You find the fun, and snap! The job’s a game.” So, we seem hard-wired to play. From each of your perspectives, what’s so important about games, or play, that a company like Microsoft would invest so much, both in research and product teams, in this area? Chris, maybe you could start us out?

Chris Bishop: Sure. From a research point-of-view, games offer an amazing environment in which to develop new machine-learning algorithms and techniques. And of course, we hope, in due course, those new algorithms will feed back, not only into gaming, but into many other domains. But beyond the, sort of, the very technical machine-learning techniques themselves, there’s also the fact that gaming is an environment in which we can explore the relationship between AI and people and see how that can work in partnership. So, it’s a very rich environment in which to drive new research ideas.

Host: Phil, what about you?

Phil Spencer: I think adding on that, what I would say is, as you said, we look at play as an innate human need. It’s something that we all do. Microsoft’s mission statement of “empower every person” not only is what we build but actually finding audience for the things that we build. So, to me, the nice thing about gaming is, it is so pervasive. People play on any device – you know, give me five minutes and a piece of paper and I’ll get bored and start playing tic tac toe with myself or something – and I think as we look at the application to empower everybody, finding human uses of technology that really can reach anybody, regardless of demographic, regardless of where you live, regardless of socioeconomic situation, I just think really opens up the opportunity for us as a company.

Host: Yeah. I have to ask, at the outset, what’s your personal relationship to gaming and why are you interested in this area, personally?

Chris Bishop: I think, yeah, my all-time greatest gaming experience was actually playing a fantastic game called Portal with my sons and really discovering that amazing social connection that it brings. It’s a very collaborative game, puzzle-solving together, and just a great father/son experience. Of course, that said, they’re a little bit more experienced than me with the Xbox controllers, so they’re kind of three orders of magnitude faster than me. But it was still good fun.

Phil Spencer: They have an unfair advantage.

Host: It’s true.

Phil Spencer: I mean, for me, I teasingly say that I think it’s the only job that Satya would hire me for at the company. I don’t think I’m kind of equipped to do anything else. But I think Chris hit right on it. Some of my best experiences with my family have been around play. I have two daughters. Some of my best friendships have been around this opportunity. I started playing with my father, he was an engineer, brought a PC home. And when I look out there today, there are so many forces in our lives that try to create division between people, the “othering” of people, you’re a this, I’m a that. And this is an art form in gaming that has a natural capability to bring people together in a cooperative social way to either win the round, to compete in a nice way with each other, and I just love that. I love that I can work in a space that has this ability to bring people together. And it’s the thing that drives me kind of personally in this space.

Host: Chris, you were on the podcast earlier this year and we didn’t talk much about gaming. Turns out you’re overseeing some pretty cool research in gaming applications, which we’ll talk about in a minute. But first, tell us why gaming is a sweet spot for the Cambridge lab and what your vision is with gaming technology research.

Chris Bishop: That’s a great question. I mean, the Cambridge lab has been around for more than twenty years and throughout that time, we’ve been very fortunate in having an incredibly close collaboration with Xbox and that’s led to a number of pieces of technology actually getting out there into the real world. And you know, one of the factors I think is just that we have a lot of enthusiastic gamers in the lab. And, you know, in research, part of what we’re supposed to do is to come up with new ideas. We’re not directed, we’re supposed to be kind of spontaneous and think a little bit out-of-the-box. And so, people naturally combine their passions. They’re excited about technology like machine-learning and many others and they are excited about games and they want to bring those two together. So, I think that’s been one of the factors for sure. We’re really, I think, uniquely placed in having this great research strength, but also being, really, the only company in that space with a, you know, a gaming division. And so, it’s just kind of an unfair advantage we have. And it’d be crazy not to take full advantage of it!

Host: Uh, Phil, you have a classic version of the mailroom-to-boardroom story here at Microsoft. You started as an intern and now you are there EVP of Gaming. Tell us how you’ve seen the company’s position in gaming evolve over the years and give us your vision as to where you see it going in the future.

Phil Spencer: Yeah, I started at Microsoft in 1988 as a programming intern from the University of Washington. Go, dawgs!

Host: Go, dawgs!

Phil Spencer: Uhhh! And you know it was interesting. Like many things that Microsoft started that were new to us back in the day, some of it was driven out of a paranoia of what competitors might be doing. I mean, Xbox was really born out of this fear that somebody else was going to create a family room PC… that we should lean in and we should go do that as a company. The thing I love about where we are in gaming today, we sit on the senior leadership team, Satya’s senior leadership team, we’re one of the key pillars of the company, is “let’s aspire to.” How do we show empathy? How do we empower? How do we use Microsoft as a platform for us in the things that we want to do? And I love that drive as a company and I just think we’ve made it through the “oh no, somebody else might do it!” to the thinking about what we might be able to do and us aspiring on our own. I’d say the partnership with Chris’ team and Microsoft Research has been so critical to that and it’s just nice to see how the company is rallying around the interest that Microsoft, all up, has in the gaming category.

Host: What’s fascinating is I’ve got a classic research person here and a classic product person and I don’t know if it’s historic or traditional separation of streams, as it were, “don’t cross the streams,” but research and product have tended to stay in their lanes. That said, I think there’s actually quite a lot of collaboration going on that people maybe don’t know about and it seems to be gaining traction now. So, do you see a sort of enhanced momentum between the academic and the industry teams and if so, why do you think that is? Chris, maybe you could start?

Chris Bishop: Yeah, I think there’s always been great collaboration, actually. It hasn’t always been so visible. But I think there is a sense of sort of more collaboration than ever. And I think partly it’s just sort of the pace of innovation. And I think there’s also more excitement in research, particularly these days, around seeing the fruits of that research really impact the real world. And, you know, one of the great privileges of being at Microsoft is you can do great research and then you can get that research out to hundreds of millions of people and that’s very tough to do that in a university environment, for example. And so, I think, you know, more than ever, there’s a real enthusiasm about very close collaboration.

Host: Yeah. How about you Phil, from the product perspective? What’s important about collaborating with these guys?

Phil Spencer: Well, we probably don’t talk about it publicly enough, but Chris’ team is our secret ingredient. When we look at the competition in our space, in gaming, the decade plus that we have working together with Microsoft Research and the Xbox team and gaming around making it more fun, making it easier to find people that you are going to have fun with, safety and security online – something we’re focusing a lot on right now – I just think it’s so critical to get people who have a sole focus, like the MSR teams do, on these categories, present a problem in areas where we’re not actually sure what the solution might be or even what it might entail, and to kind of check in every so often and see the progress that they can make. I’m really both in awe of the team and the capability that Chris has and the impact that they have on the business that we’re trying to drive.

Host: Well, let’s talk about some of the specific examples of what happens when research and product come together. I’ve had some pretty cool folks on this podcast, and they’ve brought some pretty cool research to the gaming world. Lenin Ravindranath Sivalingam has talked with me about HypeZone. And Nikunj Raghuvanshi has just been in the booth to talk about Triton. And I encourage our listeners to hit up those podcasts because it’s a deeper dive on the research. But Phil, for a second, from a product perspective, talk about how thing like HypeZone and Triton are impacting your world.

Phil Spencer: Yeah, HypeZone I think is a great example to center on. We have a streaming service called Mixer. There are hundreds of millions of people now who watch video games getting played. And I know, to some, that might seem strange. I kind of go back to the birth of like ESPN where you are going to have a television channel that only had highlights of sports and like, aaahh, that’ll never work! Things like Twitch, things like Mixer, are examples of this for the next generation of people coming up where it is their SportsCenter. It’s the place they go, the personalities they follow. So, we have this service called Mixer and it was, again, one of these “what if?” scenarios. Well, how do you know what stream to watch? How do you know where the most interesting things are happening? You know, HypeZone, we kind of borrowed the name from RedZone from the NFL of the most interesting game that’s on right now. And there’s a computer vision, machine-learning algorithm running on all of our streams across the games that have HypeZone and they’re actually looking for the most interesting moment at that time, live. And you can sit on the HypeZone channel and literally, we will swap to the most interesting instance of that game on our service. And it is about bringing the fun to somebody in an easy way. But also, when you think about the inroads that we make, and you start thinking about the applicability beyond just game streaming, well what about news on the internet? Do I follow one news service, or do I have something that knows what I like and is able to watch all the streams and bring those to me, if it’s recommendations on something like YouTube or something? I think there’s a lot of applicability for this kind of tech, which in so many areas in the things that we invest in gaming is great. I think we’re always looking at how it applies to gaming, but then also how it applies to the planet.

Host: Chris, how is it that you sort of think ahead of the curve as it were, before it hits product, what kind of research paths you’ll follow, because this could have fruit, and this doesn’t? It’s got to be difficult.

Chris Bishop: It’s a great question because research is sort of this combinatorically infinite space of possibility and the question is, you know, where should you look? And I you know often say to researchers, you know, the biggest part of their job is figuring out what research to do, and then the rest is kind of doing it. It’s almost, not the easy part, but, you know, doing the right thing at the right time. Not going after something that’s so impossible or so far out it won’t ever be real, but also just steering clear of kind of the incremental stuff, trying to do the thing which kind of sounds impossible and a couple years later you actually land it. That’s sort of the sweet spot. That stimulation of you know brainstorming about the impossible and getting ideas and thinking well, hey, maybe that isn’t so impossible. Now that kind of an idea, I could have a go at that or something. That’s actually very stimulating for researchers in terms of thinking about what they should be working on next.

(music plays)

Host: One of the biggest challenges in online gaming, and in pick-up basketball, if we’re honest, is finding players that are at your own level because everyone wants a good game, right? And apparently algorithms can help us find similar players. Talk about this new skills-ranking system called TrueSkill

Phil Spencer: Yeah.

Host: Phil, what is it and where is it being used and how is it working out?

Phil Spencer: So, I’m going to expand that a little bit because I think the fundamental issue that you talk about really, in any activity, is how do I find the group of people that I’m going to feel most comfortable with and have the most fun, or even achieve the most, if it’s something that’s more work-related. TrueSkill for us, in the pure gaming space, is like having the AI analyze how somebody plays and what their capability is, and then, when you go in to play another game, and you say I want to have a good game, I want to play against people that I’m fairly evenly-matched with, we’re able to use our understanding of your capability and put you in a match with similar skilled people. But at the same time, it’s not just the skill of the player, it’s what about what language these people use when they play? Are they aggressive players, are they not aggressive players? There’s hours of the day, even, something as simple as that, that I want to play and I don’t want to play. TrueSkill starts with your capability and builds from that, but the layering on, on top of this, of the social dynamics that happen online, that aren’t just about the skill someone has but also, when I’m there, I feel like this is an environment where I can show up as my true self. I can be most authentic and have a good time at the same time. And when I think about that, that applies in so many places in our world today, right? It’s the social interactions that we have on so many of these social networks and other things. I think the work that we’re doing together in that space is awesome. And I think there’s just so much opportunity there.

Host: Chris, could you add to that from the research perspective, because he’s talking about AI algorithms that are watching me play…

Chris Bishop: Sure. I mean, it’s a super-interesting piece of technology, TrueSkill, because you know, a lot of the work we do in machine-learning uses techniques like neural nets and deep learning and they’re very sort of data hungry and they’re amazingly powerful techniques. But TrueSkill is a really interesting showcase of a different approach to machine-learning called model-based machine-learning, or probabilistic modeling. And really, the secret there is that you, instead of having this big sort of black box that you train-up with lots of data, instead you actually describe the process by which the data comes into being. You say, well, here’s one person, they play against another person, and there’s a game outcome, and that’s the data we observe. And then you run some, you know, complicated code corresponding to some beautiful maths called Bayes’ Theorem, and it takes all the information you have about players and their skills and it absorbs the new information of the game outcome and then kind of updates their skill evaluation. And it’s all done in a way that takes into account uncertainty. So, probability is sort of at the heart of all of this. And, you know, it’s just one of the first really large-scale applications of this kind of machine learning. It went out first in Halo. It’s now part of Gears of War 4. And we’re now thinking about how we make this very generic so it can apply to a really broad range of games.

Host: Well Chris, aside from gaming, other researchers that you are working with are bringing gaming technology to healthcare. And two more guests on the podcast, Cecily Morrison, spearheaded project Torino which makes a programming language for children with visual disabilities, and Haiyan Zhang was recently on talking about several projects including Fizzyo which makes a video game out of laborious cystic fibrosis breathing exercises for kids. Talk about the vision behind the research that crosses these borders.

Chris Bishop: I mean, this is such an exciting area. I think we’re really just scratching the surface. I mean projects, like, you mentioned… we called it Torino internally, but now we’ve done a royalty-free license to an organization that’s going to produce this as Code Jumper. And it’s just a beautiful idea of allowing, initially, kids who are blind or partially-sighted, to learn to code by having, effectively, each line of code is represented as this very tactile plastic object that you connect to other plastic objects using wires and then you press a button and the code gets executed.

Host: Wow.

Chris Bishop: But actually, the really nice thing about this is that it turns out to be a great way for kids who have full vision, also, to learn to code. And it’s very inclusive, because you have the blind kids and the sighted kids can get together and work together collaboratively to design code and change code and so on. I’m just really excited that’s out there in the world now.

Host: Yeah. And what about Fizzyo? The… I mean, when Haiyan was on the podcast, I was like, that’s incredible. How did that find its way into research threads…?

Chris Bishop: You know, sometimes it’s really hard to kind of trace the sort of ancestry of these ideas. I mean, you know researchers are…

Phil Spencer: If only you could!

Chris Bishop: If only you could, yeah… I mean, research, you know, think of in a building – you got to put coffee in one end and sort of ideas come out the other and there’s this mysterious sort of organic process that happens in between.

Phil Spencer: I love that visualization.

Host: Yeah, right.

Phil Spencer: Just pour coffee in one side, innovation comes out the other. Chris’ tricks, right there!

Host: It is actually the fuel for creativity.

Chris Bishop: Absolutely. Yeah, when both coffee machines break, it’s a major…

Phil Spencer: Innovation stops!

Chris Bishop: Right? The lab director has to rush down and fix the coffee machines, otherwise…

Phil Spencer: That’s the most critical person on the team…

Host: Oh, man…

Phil Spencer: …the guy who can fix the coffee machine!

Host: Well, along those lines, Phil, millions of people saw, and were deeply moved, by Microsoft’s powerful Super Bowl ad this year about making gaming accessible for everyone. The tagline, “When everyone plays, we all win,” is perfect because it’s true: disabilities don’t reduce our desire to play. So, tell us the story behind the new adaptive controller for gamers with disabilities. Because this is another area of your oversight which is the hardware that people use to play.

Phil Spencer: Yeah, I love the story of the adaptive controller, not just because of the product that we were able to build, but really the team that started this. It started as a hackathon project inside of Microsoft, people coming together on their own time, not all from the gaming organization. Again, similar, taking a, “well, what if we tried to help in this space?” There have been a lot of third parties that had done some form of accessible controllers. Prices were very high. There were no standards. So, we took the approach if we could create a piece of hardware, basically sell it at what it cost to build, make it accessible to any other platform, so, it’s out there for Windows, it’s out there for Xbox, we’ve sent it to every other competing gaming platform saying hey, here’s a base to go use. But the team, after the hackathon project, did this in the evenings, got it to a point where they brought it to our game leadership team. We looked at the product. We weren’t sure how we were going to be able to fund it, how do you market that, what is it going to mean when it comes out? When you start that journey with that community, it’s not a one-step journey. So, if we’re in this, we’re in it for the long run. It’s not a PR beat for us, it’s a product beat. And then it was really nice, in the holidays, to see, one, when it first shipped in September, how the gaming community worked through some of the agencies that really helped us with testing and feedback to put it out. And then, there was a Microsoft ad that they put out with Owen, who is the star of the ad, great kid. And you just see him light up. But I will say, you talked about when everyone plays, we all win. The moment that gets to me the most, and I hear this so many times, is when Owen’s dad is talking about Owen playing video games. The truest representation of Owen is, when he’s playing online, he’s just Owen. He’s not in a chair. He can decide how and when he wants to share who he is. Because to everybody else, he’s a good gamer. And I love that, I think the line the dad uses is something like, you know, when he’s online playing, he’s just like everybody else. And I hear that so often in our space that the more we make it accessible, it just brings more people in, everyone can play. And something Chris hit on, which I also think is really important for us on the product side to think about, almost every instance where we’ve approached making our products more accessible, whether it’s to sight, hearing impaired people, we make the product better for everybody.

Host: Yes.

Phil Spencer: It’s actually not an either/or decision. So that it doesn’t surprise me at all that programming techniques that we would put in place for somebody who is sight-challenged becomes accessible to other people. We’ve seen that, we’ve seen that in our text-to-speech work that we’ve done on console. And it’s such a truism we have to believe, that when you make things easier for somebody, you literally make it easier for everybody.

Host: Yeah, and in fact, Cecily was talking, quite extensively, about the fact that you can have situational disabilities. You may have two arms, but you’re holding a baby and doing something with your phone. You are suddenly a one-arm phone user.

Phil Spencer: Yeah.

Host: And things like that. The curb-cut for wheelchairs is also good for cyclists. And it goes way broader.

Phil Spencer: That’s right. That’s right. It’s just expanding the aperture of how you’re thinking, and you can’t always do that because your own situation doesn’t always give you the life experience.

Host: So, Chris, on that same line, what’s the vision in terms of this accessibility thread, kind of adding on what Phil said, from your perspective?

Chris Bishop: Right. So as, I think, you know, Phil hit on something really important there which is, it’s a great frontier to drive research, take it to the next level. How are we going to get technologies to help people in social situations? And of course, that’s incredibly important if somebody is blind. So, it’s harder to get all the social cues of eye contact and that sort of thing. So, can technology help them? We’ve been doing a lot of research about how we might push the frontiers there. And then of course, it turns out that, now you take a situation like meetings, you know, video conferences, where again, like you said, we all have disabilities at different times of the day in different situations, and when somebody is on the other side of the planet and so on, there are barriers to normal communication, and then some of the research ideas apply equally in some of these situations that you might encounter in an office setting, for example.

Host: Well, some of the gaming advances we’ve talked about, and we’ve even alluded to it already here in this conversation, have come out of the annual company-wide summer hackathon…

Phil Spencer: Yeah.

Host: …but there’s a more compact version of this held for Microsoft Research and Microsoft Gaming, and it’s gained quite a bit of traction. What do you see emerging from this hack event for researchers, designers and game developers? Chris, why don’t you start?

Chris Bishop: Yeah, I think of it as sort of, you know, TED-conference-meets-hackathon. It’s a kind of a – kind of a hybrid thing we, you know, we’ve run a couple so far. The first one was quite conference-like and really sort of allowed everybody to get to know what was going on in the different teams. But the second one was really high energy and it was very focused on putting together teams from research, from gaming, and just working together on crazy little projects. I mean, one little example was some developers from the gaming teams working with researchers using machine-learning to take visuals from games and classify the genre of those games. Just a little exercise in machine-learning and a great opportunity for people to get their first experience of doing hands-on machine-learning in this very collaborative environment. But I think the real benefit that we’ve seen are the new relationships that form and people getting to know each other, and then going on to form new collaborations.

Host: So, Phil, how about from your perspective? What is this annual hack event, or now semi-annual? Biannual?

Phil Spencer: Yeah, every six months. That’s semi, right? I think… we can’t name the conference because we use a copyrighted name for the name of our little…

Host: Internally only.

Phil Spencer: That’s right, internal. Maybe I’ve seen I’ve seen it on a t-shirt somewhere, but we will not sell those. I love the fact that the teams have so much energy around this. This isn’t something that I dictated, or I think Chris dictated that must happen. It’s self-driven by the teams getting together. Many people flying in, which is time away from friends and family, to sit down and really just have interesting conversations about what opportunities are in front of us and then actually start typing and try to make some progress on certain things in a short amount of time, which is kind of the hacking nature of it. We started off with the idea we were going to do it once a year. I think, after the first one, there was this, we can’t wait twelve more months to go do it again. So, the team self-organized again to go put something in place in six months. I don’t think we’ll get to monthly. I think we’ll try to pin it at six months. But there’s just a ton of energy around those moments in time and, as we know, the work, everybody is connected now. So, that work happens even between the times where we get the teams together. But it’s literally, we have to close the doors because it fills up really quickly. But it is taking our gaming teams – not just our devs on the gaming side – our game design people, our artists, bringing them in with the researchers in categories. We usually pick a theme and they go to a secret spot and they just kind of have some fun, laugh, and also think about “what if” opportunities. And I love how it just builds on itself with the momentum the teams have.

(music plays)

Host: Chris, another researcher at Cambridge, Katja Hoffman, who was also a guest, has been spearheading some AI research on a platform called Project Malmo, which is based on the wildly popular game Minecraft, which you, Phil, I think brought into the Microsoft fold…

Phil Spencer: That’s right.

Host: …and it uses reinforcement learning to train agents. So, how is this platform enhancing the gaming experience now and how might it advance general human-AI collaboration in the future?

Chris Bishop: Right. So, Minecraft is you know, thank you Phil, that was an awesome –

Host: Awesome “get”…

Phil Spencer: Yep!

Chris Bishop: Minecraft is fantastic. One of the great things about Minecraft, it’s such a flexible environment. You can create the simplest little world up to something which is almost like the real world and everything in-between. So, from a research point-of-view, you can kind of titrate the kind of level of challenge that you are presenting your AI. If it’s too easy, just make it harder. And so last year, we used Minecraft, under this Malmo umbrella, to launch a competition to allow researchers around the world to experiment with AI in the domain of reinforcement learning. And it’s been very, because of this sort of open nature and collaborative nature, it’s just brought lots and lots of people together. In fact, even within the company, I mean, it’s been a great collaboration amongst the different MSR labs. It’s not just Cambridge. And Katja does a, you know, fantastic job of leading this. But you know, our labs in Redmond and Montreal, and just bringing researchers across Microsoft internationally together along with the great folks at Minecraft.

Host: So, I want to shift a little over to the fact that we live right now in a fairly console-centric world, but I’m told the future of gaming is anywhere, with anyone, on any device. So, what role does cloud computing play in the gaming world? What can it do for both gamers and game developers?

Phil Spencer: Yeah, in today’s world, there’s a little over seven billion people on the planet. About four billion of those are connected to the internet in some way, whether it’s a mobile phone, a PC, a console, any other device. Half the connected world plays video games today.

Host: Amazing…

Phil Spencer: And, if you think, by 2030, analysts are saying four out of the five billion connected consumers at that point will be playing games. Not everybody considers themselves a gamer, but you give them twenty minutes on a bus with a phone, they’re probably going to start playing something. So, when we think about our ambition as Microsoft, what scale should we think about? We don’t think it should be about what device you own. We put the player at the center of the decisions that we make, and we say, how do we bring the game’s content that you want to you on any device? How does your community move with you from device to device? And this happens on any other – you’re a Netflix customer on your PC, on your phone, on your television. Anywhere you go, you can log in and use Spotify, Netflix, any service, Facebook, Twitter, all of these things meet you where you are. Gaming is kind of the last bastion of device-first consumer engagement. And as Microsoft, we say we won’t accept that. We want to make you the owner of your experience and you should be able to go from device to device. The issue we have is, not all the devices are technically capable of displaying some of the rich worlds and characters and stories that our creators create. So, cloud allows us to do the rendering of that experience in our datacenters and then bring the video stream down, literally, to any device that can decode a video and take some input and allow people to play anywhere. And I always think this isn’t really about people who say, okay, I’m going to play while I’m away from my PC or away from my console. Like yeah, those will be users. But I think about people in emerging markets where their only compute device will be the phone in their pocket. But the connection of the internet means they know what a Halo is, right? They know what a Minecraft is. They know what our brands and what our stories and characters are. And we’re pushing to bring those experiences to the devices that they already have in their pocket.

Host: So, Nikunj was just talking about the “secret sauce” of pre-computing and having Azure manage that because it’s just so expensive, in CPU time, to do these big calculations that you need for realistic games. So, from your perspective Chris, how does the cloud help out the work you are doing upstream of the device delivery?

Chris Bishop: Yeah, the cloud is such an amazing innovation, sort of the elasticity of computation, the ability to effectively put super computers, super-super-hyper-computer capabilities, you know, in the pockets, as Phil said, of somebody of Africa, effectively. And this was sort of inconceivable in the pre-cloud era. So, that scale of computation is incredibly important in the backend of research because it’s driving, really, the state-of-the-art in machine learning and AI. It’s also really important for these, sort of, rich simulated worlds. I mean, one of the great technologies that comes out of the gaming world is very rich simulated worlds which can be used for other kinds of things as well. Reinforcement learning for, you know, autonomous vehicles or whatever it may be, or applications in healthcare and so on. So, you know, the cloud really plays a very fundamental role in the AI revolution, very generally, I think.

Host: So, the mission of Microsoft is to, quote-unquote, empower every person and organization on the planet to achieve more. And it’s an interesting context putting it in the gaming world, how do I achieve more in gaming? I win more, I get a better social experience, there’s all kinds of interesting threads on that that you guys have addressed that I actually never thought of, which was fantastic. How is gaming, and the technical research behind it, important to the company’s mission, and what is it about gaming and gaming research that empowers people?

Chris Bishop: One really interesting thing about AI in gaming, from a Microsoft perspective, is that it’s not about how do we get a computer to beat the human, right? How do we achieve human-level performance in game X, game Y, game Z, as if a human being were somehow a list of capabilities and you’re crossing them off, one at a time? This is, you know, this is a sort of much a richer view of AI, as a collaboration between people and machines, effectively. How can the AI enhance the human experience, in this case, you know, create more fun, create more rich, immersive, interesting, varied gaming experiences? So, in terms of empowering people, it’s really part of that AI narrative of the complementarity between AI and people. And it’s true today, and I think it will be true for a very long to come, that the capabilities of the machine and the capabilities of people will be different and they will be complementary and they come together in this very nice synthesis in the world of gaming.

Host: Phil, talk to us about your side.

Phil Spencer: Yeah, I really think about the word empower and what does that mean to a person? And I don’t think it’s Microsoft’s job to define what “empower” means. I think about our work in the gaming space. It can be messy at times, like the world can be messy. There’s violence. There’s people who play too often. There’s the monetization, how that works. One of the easy things that we can do as a company is lean back and say okay, it’s too messy, we don’t want to play in that space. But then when I look at our mission statement of empowering every person, empowering every person’s also going to be a little messy every once in a while. And I love that we learn and get the feedback through the activity of being relevant in a space where people are gravitating. I think, how many kids on the planet, their first compute experience is picking up mom or dad’s iPad and playing Minecraft? And what responsibility do we have, as the curators of Minecraft, in that loop with somebody? The social interactions they are going to have when somebody comes into their world, what should that feel like? How does a five-year old feel empowered in that situation? And what do we teach, both ourselves and the people in that environment, at those influential times that will permeate how people work, how they interact with each other in different social environments, whether it’s physical or digital? I love the opportunity to empower, and I just think the gaming space gives us this opportunity to empower people to achieve more, and maybe the “achieve” is kind of satisfaction and fun, but also to tackle some of the issues that are out there that are Microsoft scale opportunities. And the interaction right now between the company’s values and principles in this space and what we want to stand for in the gaming space, I don’t think have ever been stronger. And it’s… I think it’s just an amazing opportunity for us.

Host: So, I ask all the guests that come into this booth some version of the question “what keeps you up at night?” So, what should we be thinking about in the worlds that you are responsible for creating and shipping? What keeps you up at night?

Phil Spencer: I’m a go-to-bed-at-ten person… I’m not a high-stressed person, I just don’t work that way. But I like the question, and when I think about, as the head of gaming at Microsoft, there are two sides to anything that you do. What content should we approve? What activities on the platform are safe for others? How long should somebody play? Who gets to control that? How they spend? As I said earlier, I want to be part of that conversation, not outside of it. If we decided that we wouldn’t want to do gaming anymore, gaming’s not going away. So, the thing that motivates me is I’d rather be at the table and part of the conversation than just sitting back and pointing at it. Because, as we started at the beginning of all this, play is an innate human need. People are going to play. And I think we have a unique opportunity to be part of that. And part of that means dealing with the hard issues and the sticky issues and the things that maybe keep us up at night, sometimes. But I’d rather be relevant in that conversation and part of it and bring a Microsoft point-of-view, than outside of it just pointing at it as something that’s too hard for us. I won’t accept that.

Host: Chris, what do you think? What keeps you up at night?

Chris Bishop: I’m what you call a technology optimist, right? I really believe in the amazing power of technology to do all kinds of great things in the world. And in particular, technologies like AI and really machine-learning, they’re ubiquitous. They will be used in many, many different places. But when you have a very powerful general technology as we know, there are ways that it could be, you know, deliberately misused, but also there are just ways in which you could accidentally kind of get it wrong. And we’ve even seen in the, you know, the early phases of this revolution, situations, not too much in gaming, but areas where people, very well-intentioned people, have hit some little bumps in the road, perhaps around biases and just unintended consequences of learning from data. But I think the good news here is that we, you know, particularly Microsoft as a company, but I think we as a community, are very much on the front foot. We’re recognizing that there are these obstacles, these challenges and, in many cases, they actually become research problems, and they’re very actively being researched. So again, I think, once we recognize that we can be on the front foot and really try to avoid most of those.

Host: Well this has been a fascinating conversation and it’s not quite over yet because I have one more question. But I want to thank you guys both for this eye-opening, Venn diagram overlap descriptions of this world. So, here’s your chance to talk to would-be researchers and developers who are also gaming enthusiasts out there. Um, I love this quote from Satya. He says, “We don’t want to be the cool company in the tech sector, we want to be the company that makes other people cool.” I don’t know if he actually said that, but if he didn’t, he should have. What’s on the horizon for the future of gaming and why is Microsoft a good place for the cool kids to be?

Chris Bishop: So, I actually think of three things here that I’d just love to see come out of this. I mean, the first one is seeing this amazing collaboration really driving the frontiers of AI. We’re seeing this already, new algorithms, new techniques for machine-learning being driven by the challenges that arise in the world of gaming. So, that’s the first one. And then the second one is to see the flow of ideas in the other direction, to see that we’re really impacting the world of gaming, creating rich new experiences and just do wonderful things for hundreds of millions of people, if not billions of people, around the world. That’s an amazing opportunity. And then even the third one is even more ambitious, to see this amazing collaboration lead to things that spill over and spin out into other fields, and the healthcare one is a particularly exciting one. So, you know, we’re really quite ambitious!

Host: Phil, how about you?

Phil Spencer: Yeah, I think that maybe we can be the cool kids and make others cool at the same time? But the opportunity to put our customer, whether the customer is a developer or the people using our products at the center is really what I hear from Satya. It’s not about our gratification in the work that we do, it’s what impact does that have on others? I think that is such a pivot for us. Microsoft, for a long time, has been centered on what we do and maybe how we do it. Now we’re getting more and more focused on why we do what we do. But I love that transformation that I’ve seen inside the company. And I think so much of the work that’s going on right now just has application that we see today, but also application that we don’t see tomorrow. And that’s why I think the ongoing work for all these developers, on Lean In and Learn, growth mindset. I can kind of take off our cultural principles. But I think that we have such a rich place to go learn from and some incredible people to learn from, like Chris and Haiyan and the team at MSR is just awesome, and I love that opportunity. It’s such a great thing about working here.

Chris Bishop: And it’s just a great chance to say a huge “thank you” to Phil and your amazing team. And I, you know, it’s been a fantastic collaboration. I’d really just love to see this carry-on for many years to come.

Phil Spencer: Great. Thanks, Chris.

Host: And as it turns out, cool is not a zero-sum game. We can be cool, you can be cool, we can all be cool.

Phil Spencer: That’s right. Cool in our own way!

Host: And you guys – you guys are really cool. So, thanks to both of you, Chris and Phil, for making time to share your insights on the podcast today. Great to have you both here in person.

Chris Bishop: Thanks.

Phil Spencer: Thanks for the questions.

(music plays)

To learn more about Dr. Chris Bishop and Phil Spencer, and the how Microsoft takes playing games seriously, visit Microsoft.com/research

Up Next

chris bishop

Artificial intelligence

Machine learning and the learning machine with Dr. Christopher Bishop

Episode 52, November 28, 2018 - Dr. Christopher Bishop talks about the past, present and future of AI research, explains the No Free Lunch Theorem, talks about the modern view of machine learning (or how he learned to stop worrying and love uncertainty), and tells how the real excitement in the next few years will be the growth in our ability to create new technologies not by programming machines but by teaching them to learn.

Microsoft blog editor

Katja Hofmann

Artificial intelligence, Search and information retrieval

Malmo, Minecraft and machine learning with Dr. Katja Hofmann

Episode 39, August 29, 2018 - Dr. Hofmann talks about her vision of a future where machines learn to collaborate with people and empower them to help solve complex, real-world problems. She also shares the story of how her early years in East Germany, behind the Iron Curtain, shaped her both personally and professionally, and ultimately facilitated a creative, exploratory mindset about computing that informs her work to this day.

Microsoft blog editor

Artificial intelligence, Computer vision

Machine learning and the incredible flying robot with Dr. Ashish Kapoor

Episode 22, May 2, 2018 - Dr. Kapoor talks about how cutting-edge machine learning techniques are empowering a new generation of autonomous vehicles, and tells us all about AirSim, an innovative platform that’s helping bridge the simulator-to-reality gap, paving the way for safer, more robust real-world AI systems of all kinds.

Microsoft blog editor